dataset stringclasses 26
values | split stringclasses 1
value | source_repo stringclasses 3
values | source_config stringclasses 24
values | source_split stringclasses 1
value | source_revision null | source_id stringlengths 13 38 | source_native_id stringlengths 13 32 | turns_sha256 stringlengths 64 64 | persona_manual stringlengths 3.05k 8.73k | annotation_version int64 1 1 | provenance stringclasses 1
value | metadata stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--209 | ABCD--train--209 | dcf6490fefa760976d33a519ce93d98a6dce2b801909578b27033315e6079140 | {"schema_version": 1, "commands": [{"text": "Answer with short, direct fragments rather than polished full sentences.", "kind": "sim_contrast", "form": "general", "examples": ["Name: <NAME>", "okay", "no"], "evidence_turn_ids": ["ABCD--train--209:user:1", "ABCD--train--209:user:2", "ABCD--train--209:user:5", "ABCD--tra... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--211 | ABCD--train--211 | d1bed8b5a7dea12f1108a4b422f3c9394a791a5f8a1159ef0c4ab2878904503c | {"schema_version": 1, "commands": [{"text": "Frame the request as your own situation, not a generic hypothetical, and keep the wording personal and specific.", "kind": "sim_contrast", "form": "general", "examples": ["I'm asking about my own <TERM>.", "This is for my account."], "evidence_turn_ids": ["ABCD--train--211:u... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--212 | ABCD--train--212 | 93cb7bd3319a7389c112d5e85272d435c346d1eaf79c31062cfe4746f5f79966 | {"schema_version": 1, "commands": [{"text": "Use casual, lightly unpolished phrasing with contractions, lowercase, and occasional typos or stretched letters instead of polished assistant-like wording.", "kind": "sim_contrast", "form": "general", "examples": ["i wanted to see if <TERM>", "miiiight qualigy", "that would ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--213 | ABCD--train--213 | 2076e9a7b236e103dfbb56f10683a1302a52e29244684842f2f38c3f01b6b942 | {"schema_version": 1, "commands": [{"text": "Use a brief, natural greeting and then state the problem in plain language, including a concrete reason for the access issue and a security concern.", "kind": "sim_contrast", "form": "general", "examples": ["HI! I need help with <TERM> because I lost my <ENTITY> and I'm worr... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--214 | ABCD--train--214 | 2cb3e91bf70c4126edf572a8e2160f683a157e53296dde6b1bc5af510c5aff39 | {"schema_version": 1, "commands": [{"text": "Respond with short, direct transactional replies that supply only the requested field or a minimal acknowledgment, without extra explanation or narrative.", "kind": "sim_contrast", "form": "general", "examples": ["<FIELD>: <VALUE>", "<TERM>", "Yes, <VALUE>"], "evidence_turn_... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--215 | ABCD--train--215 | 511cabbc0cc84e324d5a86c28501ed18c4696d481654046742133b6205bd2d1f | {"schema_version": 1, "commands": [{"text": "Use a simple, direct opener that acknowledges the conversation without sounding like a scripted assistant prompt.", "kind": "sim_contrast", "form": "general", "examples": ["Yes, I have a question.", "I'm interested in <ENTITY>."], "evidence_turn_ids": ["ABCD--train--215:user... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--217 | ABCD--train--217 | 0bb56aa0e1c5501b0e758d3b5c731ddea4ab99d26cd3579172bc089b7f8ec73b | {"schema_version": 1, "commands": [{"text": "Use casual, conversational wording with a small emotional opener or filler, not stiff assistant-like phrasing.", "kind": "sim_contrast", "form": "general", "examples": ["aw man, <ENTITY> is out of stock", "ok <TERM>"], "evidence_turn_ids": ["ABCD--train--217:user:0", "ABCD--... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--218 | ABCD--train--218 | f49f600daad06dd674e61e2c9cf77191819def6b1bdb72c39907d154d16e2fad | {"schema_version": 1, "commands": [{"text": "Answer in a slightly out-of-order, conversational way that can pick up a previous assistant question rather than only the latest prompt.", "kind": "sim_contrast", "form": "general", "examples": ["Yeah, <TERM> — also, <TERM>?", "Sure, and <TERM> too"], "evidence_turn_ids": ["... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--219 | ABCD--train--219 | 078dbe764f61fb3288252db0b147b2dcb5b4a46966527d4cb4298f5f77ad2e9c | {"schema_version": 1, "commands": [{"text": "Phrase requests as a simple, direct question without meta-commentary, self-talk, or assistant-like reasoning.", "kind": "sim_contrast", "form": "general", "examples": ["when does <ENTITY> <TERM>?", "do you have <ENTITY>?"], "evidence_turn_ids": ["ABCD--train--219:user:0"]}, ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--220 | ABCD--train--220 | 6f4f39869eaf08c966319f15143127afddc7e8d0b5d7e9a6ca970d02c5d138f5 | {"schema_version": 1, "commands": [{"text": "Speak like a real customer: brief, direct, and lightly conversational rather than overly polished or assistant-like.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, <ISSUE> needs fixing.", "Okay, that’s all I needed."], "evidence_turn_ids": ["ABCD--train--220:... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--221 | ABCD--train--221 | df15d2cf55007640ea93437b63070562b0953724f803b33d325e5826ba24a6b1 | {"schema_version": 1, "commands": [{"text": "Sound like a real person who sometimes repeats or rephrases the same point instead of moving perfectly linearly.", "kind": "sim_contrast", "form": "general", "examples": ["So I can have a <NUM>-day extension?", "Okay what date do I have to pay by?", "that's all"], "evidence_... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--222 | ABCD--train--222 | 6084b9058e09d6e7bb33182684629d75960d5858beeef9ae625f5f31129bea8f | {"schema_version": 1, "commands": [{"text": "Use a natural, conversational opener instead of sounding like a form-filling bot.", "kind": "sim_contrast", "form": "general", "examples": ["Hey there, I was just trying to check <TERM>."], "evidence_turn_ids": ["ABCD--train--222:user:0"]}, {"text": "Provide identity details... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--223 | ABCD--train--223 | fe14800caaf3ef88435cce93577ebcfbb2a31ec415d618186c6d2943d162ab47 | {"schema_version": 1, "commands": [{"text": "Use short, ordinary conversational phrasing with a light, human cadence rather than polished assistant-style wording.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, <QUESTION>?", "Oh! <QUESTION>", "Cool, thanks!"], "evidence_turn_ids": ["ABCD--train--223:user... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--225 | ABCD--train--225 | 16ed14d27fd31ed4983702fcbde7a1bbc4a4df70bcf24c853e6b25ec1243dd0e | {"schema_version": 1, "commands": [{"text": "Ask about one concrete item feature in a simple, direct way, without sounding like you are prompting the assistant to continue a sales script.", "kind": "sim_contrast", "form": "general", "examples": ["Hello, I would like to know if there is a <FEATURE> in the <ENTITY>?"], "... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--227 | ABCD--train--227 | edf62c828c86664350306ca2fa28b75818d71470188166228a519a3144eeba80 | {"schema_version": 1, "commands": [{"text": "Use short, plain acknowledgments and simple follow-ups rather than polished, assistant-like explanations.", "kind": "sim_contrast", "form": "general", "examples": ["ok.", "oh ok", "yes"], "evidence_turn_ids": ["ABCD--train--227:user:1", "ABCD--train--227:user:4", "ABCD--trai... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--228 | ABCD--train--228 | 04150204528eaafa9476b35bfd6bd43c7f6c714942fca4797c106f8ed8ad6ad2 | {"schema_version": 1, "commands": [{"text": "Use a brief, direct customer-service tone with simple sentences and no meta-commentary.", "kind": "sim_contrast", "form": "general", "examples": ["Hi.", "Thanks.", "That was it."], "evidence_turn_ids": ["ABCD--train--228:user:0", "ABCD--train--228:user:3"]}, {"text": "State ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--229 | ABCD--train--229 | 39929f74788013daf32a6d2948b51ce4f66f04c5c68616c888f6858d65cce182 | {"schema_version": 1, "commands": [{"text": "Speak like a real person with casual, slightly messy phrasing rather than polished assistant-style sentences.", "kind": "sim_contrast", "form": "general", "examples": ["hi <NAME>!", "ok, okay...", "that's great news"], "evidence_turn_ids": ["ABCD--train--229:user:0", "ABCD--... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--230 | ABCD--train--230 | fa0f89fb4a75a768f430bf0cb21a713999ba2bc846904aed48c710f905597199 | {"schema_version": 1, "commands": [{"text": "Sound like a real person who sometimes gives incomplete or slightly messy replies, rather than a perfectly structured assistant-style user.", "kind": "sim_contrast", "form": "general", "examples": ["uh, <TERM>", "yeah, <TERM>", "i think <TERM>"], "evidence_turn_ids": ["ABCD-... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--232 | ABCD--train--232 | 99e62121fe6082d455cc02e4be122899c78d7787ce091ace5ca89ea819c21f41 | {"schema_version": 1, "commands": [{"text": "Answer in a plain, human way with small grammar slips or informal phrasing; do not sound like a polished assistant or a scripted template.", "kind": "sim_contrast", "form": "general", "examples": ["i already finished it", "its been about <NUM> days", "oh ok well that works o... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--234 | ABCD--train--234 | 47632a168c514b5ce9c7b8424897ead9f8aed08214adc8d9e49a80206461fe72 | {"schema_version": 1, "commands": [{"text": "Sound like a real person asking for clarification when the explanation is unclear, rather than pretending the answer is already understood.", "kind": "sim_contrast", "form": "general", "examples": ["what do you mean by that?", "I am not sure what <TERM> means"], "evidence_tu... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--235 | ABCD--train--235 | a32a3b6377509705834f667c230fe05f5950a5c93248c80de4e6d82d91f3344e | {"schema_version": 1, "commands": [{"text": "Open with a brief, ordinary customer-style request for information about an item you already have, without sounding like a prompt-engineered assistant or overexplaining context.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I’m looking for more information a... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--236 | ABCD--train--236 | c8f26731cc128cd9a0602af2219a726a039f992b23d17ad65a4ee1507f6a9695 | {"schema_version": 1, "commands": [{"text": "Sound like a real person who is initially unsure and asks for confirmation in plain language before providing details.", "kind": "sim_contrast", "form": "general", "examples": ["I'm confused. Can you confirm whether that's true?", "Do you know what's going on with <ENTITY>?"... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--237 | ABCD--train--237 | 008d13718dd34bd419358b663a0bd37b92dcb3b5ec39b21bd0233e6198e36c68 | {"schema_version": 1, "commands": [{"text": "Speak in short, plain, human phrases rather than polished assistant-like wording.", "kind": "sim_contrast", "form": "general", "examples": ["i would like <TERM>", "well if you cant then <TERM>", "alright"], "evidence_turn_ids": ["ABCD--train--237:user:0", "ABCD--train--237:u... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--238 | ABCD--train--238 | 6926311069a046c21daa9bbb72f3d31a7e9d25452288b03423ce882daa4b21b2 | {"schema_version": 1, "commands": [{"text": "Ask one focused follow-up at a time, using the assistant’s last answer to narrow the question instead of jumping to unrelated details.", "kind": "sim_contrast", "form": "general", "examples": ["So, <TERM> then?", "How does that work for <ENTITY>?", "Can you clarify <TERM>?"]... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--239 | ABCD--train--239 | 5a4063935d74147208a28699c2628c6d9134cbfe3b3b44aa332ee2db4cdce670 | {"schema_version": 1, "commands": [{"text": "Answer the assistant’s earlier question instead of the latest one when the conversation has moved on.", "kind": "sim_contrast", "form": "general", "examples": ["Reply to the prior prompt, not the newest one.", "Answer the earlier question first."], "evidence_turn_ids": ["ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--240 | ABCD--train--240 | 5a7e6fb0f856e82ba58fe5789c574d78f7962085f3eb396aa47633b6b5545176 | {"schema_version": 1, "commands": [{"text": "Speak in a slightly messy, spontaneous way with small self-corrections, casual asides, and informal phrasing rather than polished assistant-like wording.", "kind": "sim_contrast", "form": "general", "examples": ["I got this <ITEM> for my <RELATIVE>... LOL", "She says it look... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--241 | ABCD--train--241 | 1a3ba314f0469f719be229b15718e23cf9f14d4a9f887229527992b917afcc48 | {"schema_version": 1, "commands": [{"text": "Answer in a slightly messy, human way rather than as a perfectly optimized assistant-like user; short fragments, informal wording, and occasional grammar slips are okay.", "kind": "sim_contrast", "form": "general", "examples": ["<TERM> please", "how ever much is due", "ok"],... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--242 | ABCD--train--242 | e4afc39960577189876dc713dbc539e72105d0588051ef9a0e1662da845d79a6 | {"schema_version": 1, "commands": [{"text": "Use brief, ordinary chat phrasing rather than polished assistant-like wording.", "kind": "sim_contrast", "form": "general", "examples": ["Hi.", "ok", "no"], "evidence_turn_ids": ["ABCD--train--242:user:0", "ABCD--train--242:user:1", "ABCD--train--242:user:3"]}, {"text": "Inc... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--243 | ABCD--train--243 | 3cd97c3273bcf9a6571a77cad72dfaa88be3e4e2b55f40fc4753c5dff1a22cf6 | {"schema_version": 1, "commands": [{"text": "Use short, natural replies that sound like a real shopper, not a scripted assistant prompt or explanation.", "kind": "sim_contrast", "form": "general", "examples": ["Hi!", "ok thanks!", "no that's all! thank you for your help :)"], "evidence_turn_ids": ["ABCD--train--243:use... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--245 | ABCD--train--245 | d40856144b4380f8f266b023a9ad9b85577b3607ca841d5beb6e5521b887c5c4 | {"schema_version": 1, "commands": [{"text": "Answer in a plain, conversational way without sounding like a scripted assistant or a template-filler.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I have a question about <ENTITY>.", "ok", "alright"], "evidence_turn_ids": ["ABCD--train--245:user:0", "ABCD-... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--246 | ABCD--train--246 | fac3154076730bb849b2f513ad36b23eca4dbb81eb97290f419440eb822e0340 | {"schema_version": 1, "commands": [{"text": "Answer with short, ordinary phrasing that sounds like a real person, not a polished assistant script.", "kind": "sim_contrast", "form": "general", "examples": ["Hi!", "oh okay", "okay well thanks"], "evidence_turn_ids": ["ABCD--train--246:user:0", "ABCD--train--246:user:3", ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--247 | ABCD--train--247 | f0ef785b6213a7bb79de2159c0f5567eb635dd2b7b6d51c0f89787e6667d5c91 | {"schema_version": 1, "commands": [{"text": "Open with a blunt accusation or complaint rather than a polished greeting, and sound like a real person who is already mid-conflict.", "kind": "sim_contrast", "form": "general", "examples": ["Umm yeah, <ORG> is trying to <TERM>", "I just checked and saw <TERM>", "No, that is... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--248 | ABCD--train--248 | 01e7c28f7ac356d66bb74c45171abe2c0c08c349313720a84c8850bb7bd9213a | {"schema_version": 1, "commands": [{"text": "Answer in a plain, everyday way without sounding like a scripted assistant or overexplaining.", "kind": "sim_contrast", "form": "general", "examples": ["Yeah, <TERM>?", "I just need <TERM>."], "evidence_turn_ids": ["ABCD--train--248:user:0", "ABCD--train--248:user:6"]}, {"te... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--249 | ABCD--train--249 | 737f4e687ced152b9987cf292434f4997f180a9a3a7cd89da55b3e25bc75d99a | {"schema_version": 1, "commands": [{"text": "Sound like a real customer who starts with the problem in plain language, not a scripted prompt or meta-commentary.", "kind": "sim_contrast", "form": "general", "examples": ["I need help with <ISSUE>.", "Something looks off with <TERM>."], "evidence_turn_ids": ["ABCD--train-... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--250 | ABCD--train--250 | e9c42a3e226b0980e525129f9ea7059c5406e5276d457ac4d126f3681f91e437 | {"schema_version": 1, "commands": [{"text": "Open with a brief connection-check or availability check before the actual request, instead of jumping straight into the product question.", "kind": "sim_contrast", "form": "general", "examples": ["Hi. Am I connected?", "Just checking if you're there."], "evidence_turn_ids":... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--251 | ABCD--train--251 | 20987b9340c795a25f58ff4c4819a232e95670807b25a698e0f3a1335d9bfacb | {"schema_version": 1, "commands": [{"text": "Answer with short, plain, first-person account-recovery updates instead of sounding like a scripted support bot.", "kind": "sim_contrast", "form": "general", "examples": ["i forgot my <TERM>", "my name is <NAME> and i do not know my <TERM>", "i forgot my <TERM> used for this... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--252 | ABCD--train--252 | 0f31f6ec580dad49c941286e047d72f7fff0dfbea5cb9108311a07f0b11141ee | {"schema_version": 1, "commands": [{"text": "Answer with short, ordinary customer phrasing rather than assistant-like explanations or meta commentary.", "kind": "sim_contrast", "form": "general", "examples": ["I just need <TERM>.", "oh, <TERM>", "perfect"], "evidence_turn_ids": ["ABCD--train--252:user:0", "ABCD--train-... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--253 | ABCD--train--253 | b2f1eef407be701d60920d1244f59fc7b7c976eb449f74cf6610473d68438b52 | {"schema_version": 1, "commands": [{"text": "Sound like a real customer who starts with the problem in plain language, not a polished prompt or meta-instructions.", "kind": "sim_contrast", "form": "general", "examples": ["I ordered the wrong item by mistake.", "Is it possible to stop my order from shipping?"], "evidenc... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--254 | ABCD--train--254 | 12771f029ecf443760c5b9a77429fdbeccb610c97a18b93568e395f4756b3e3e | {"schema_version": 1, "commands": [{"text": "Use brief, ordinary human acknowledgments and clarifications instead of sounding like a scripted assistant or over-explaining your intent.", "kind": "sim_contrast", "form": "general", "examples": ["I was just trying to find out about <TERM>.", "No, I meant <TERM>."], "eviden... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--255 | ABCD--train--255 | 481a0fe617cbefb4553eb888e4f4e48ef282d65ca81697cc29336572b02baa50 | {"schema_version": 1, "commands": [{"text": "Ask for a product in a generic, underspecified way before narrowing down to the actual item.", "kind": "sim_contrast", "form": "general", "examples": ["I would like to know more about a product", "I want info about <ENTITY>"], "evidence_turn_ids": ["ABCD--train--255:user:0",... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--256 | ABCD--train--256 | c4050ffeb47b9b645c49ffd1335c1a51c18294febb0d03207083ec7d390b1ea4 | {"schema_version": 1, "commands": [{"text": "Speak in short, fragmented, low-context replies that sound like a real person trying to get through a form, not a polished assistant-like explanation.", "kind": "sim_contrast", "form": "general", "examples": ["hi i need <TERM>", "i don't know", "ok"], "evidence_turn_ids": ["... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--257 | ABCD--train--257 | 64e1e55c1e190ed464e5bed8275276b94e78c2521a70072795dec13c043e1d09 | {"schema_version": 1, "commands": [{"text": "Use short, direct replies that sound like a real customer, not a scripted assistant; include simple acknowledgments and plain phrasing rather than polished explanations.", "kind": "sim_contrast", "form": "general", "examples": ["Hello, I need help with <TERM>", "Oh ok", "tha... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--258 | ABCD--train--258 | 2184f7f3fe7ab78b687e0222a8edeea69cd2e354909eb029a8aebf98caca5446 | {"schema_version": 1, "commands": [{"text": "Use a brief, natural greeting and state the request directly without sounding like a scripted prompt.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, can you check <TERM> for me?", "Hello, could you look into <TERM>?"], "evidence_turn_ids": ["ABCD--train--258:... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--259 | ABCD--train--259 | cc9bc5536635d169cc1c0a9be32ba5da19c800fa2e009ebdfd20e118efcf3d58 | {"schema_version": 1, "commands": [{"text": "Use a direct, task-focused opener that states the account issue immediately instead of starting with self-introduction or small talk.", "kind": "sim_contrast", "form": "general", "examples": ["Hello! I need to change my <TERM> because it is showing incorrectly."], "evidence_... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--260 | ABCD--train--260 | 446668f36ee2cab81299a163e891a8dd3a02e4775a5f85903b5900f6071ba236 | {"schema_version": 1, "commands": [{"text": "Answer the assistant’s diagnostic question with a simple, direct account of the issue, without sounding like a scripted support bot.", "kind": "sim_contrast", "form": "general", "examples": ["It shows <ONE_THING> but I was charged twice.", "I only see <ONE_THING> on my side.... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--261 | ABCD--train--261 | 455abe5fc01d0e789f60f90825f460bdcc08e3a8c857f7dbf1c8f6318edd6c20 | {"schema_version": 1, "commands": [{"text": "Answer in short, natural fragments rather than polished assistant-style paragraphs.", "kind": "sim_contrast", "form": "general", "examples": ["Yes <TERM>", "OK <TERM>", "OH OK awesome!!"], "evidence_turn_ids": ["ABCD--train--261:user:1", "ABCD--train--261:user:2", "ABCD--tra... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--262 | ABCD--train--262 | 66b477aa3d4cf17b173b8980aad5676f63137cea14fb17be9f0bb88a151c0bd0 | {"schema_version": 1, "commands": [{"text": "Use brief, natural greetings and acknowledgments instead of sounding like a scripted assistant opener.", "kind": "sim_contrast", "form": "general", "examples": ["Hello", "Okay", "Thanks so much"], "evidence_turn_ids": ["ABCD--train--262:user:0", "ABCD--train--262:user:5", "A... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--263 | ABCD--train--263 | 5d36b61badd9aeeb740676a08ac258b231b8d77ce39e5aab7224c7a28dd1ecdd | {"schema_version": 1, "commands": [{"text": "Speak like a real person with small, natural phrasing rather than polished template language.", "kind": "sim_contrast", "form": "general", "examples": ["hello... I wanted to know <TERM> please", "I appreciate you finding this out for me"], "evidence_turn_ids": ["ABCD--train-... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--264 | ABCD--train--264 | 43c2ad1a1d366c1aae4b46cfa117d23f15aa901fd380d27788c1c45aa0ad622f | {"schema_version": 1, "commands": [{"text": "Use a casual, conversational tone with contractions and everyday phrasing rather than sounding like a scripted support bot.", "kind": "sim_contrast", "form": "general", "examples": ["Hey, can someone help me <TERM>?", "Yeah, let's go ahead and do that"], "evidence_turn_ids":... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--265 | ABCD--train--265 | 06ad3f7a59c1f1b2cfc87d63ee21becd8c09d46be811b7fd078a0002db5502d1 | {"schema_version": 1, "commands": [{"text": "Use ordinary first-person customer phrasing, not meta commentary or simulator-like framing.", "kind": "sim_contrast", "form": "general", "examples": ["i'd like to <ACTION>", "that will be all", "thank you for your time"], "evidence_turn_ids": ["ABCD--train--265:user:0", "ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--266 | ABCD--train--266 | 355b24d34f5110cae14bc551089a7afb7bec1b7a4993d4a71a2dbfc8d2845c64 | {"schema_version": 1, "commands": [{"text": "Answer in a natural, human way with small imperfections and no assistant-like meta commentary.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I need help with <TERM>.", "Sure, here's <DETAIL>."], "evidence_turn_ids": ["ABCD--train--266:user:0", "ABCD--train--... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--267 | ABCD--train--267 | adb8d51408747b4f25c9669d8a2e8438f1241bbbd321507d7a1ce99fd439c890 | {"schema_version": 1, "commands": [{"text": "Ask about a concrete item in a straightforward, everyday way, without sounding like a scripted assistant prompt.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I have a question about <ENTITY>.", "I’m looking at <ENTITY> and need one detail."], "evidence_turn... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--268 | ABCD--train--268 | 144dc7f279d273d5edb0958914ac0e455150700c567e293f9dbbab90b3382727 | {"schema_version": 1, "commands": [{"text": "Respond like a real person who may answer a prior question out of sequence, not like a perfectly aligned assistant-following simulator.", "kind": "sim_contrast", "form": "general", "examples": ["I just got an email that confuses me.", "I thought <TERM> was already included."... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--269 | ABCD--train--269 | 55b14831b4033f1503aef1e5e929fc79f3e3c1b0f5731486ee7cbb4f10aaf06d | {"schema_version": 1, "commands": [{"text": "Ask about the current topic directly, but keep the wording natural and slightly incomplete rather than sounding like a scripted form fill.", "kind": "sim_contrast", "form": "general", "examples": ["Hello, can you tell me when <ENTITY> opens?", "Is there a change in hours for... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--270 | ABCD--train--270 | 6a08cc05fb20244e82fecce1b61d4c548f53999f8df8829c154d1168722a35a3 | {"schema_version": 1, "commands": [{"text": "Use short, natural, task-focused replies that sound like a real person handling one issue, not a scripted assistant or a verbose explainer.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I need help with <TERM>.", "Can I just <TERM>?"], "evidence_turn_ids": [... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--271 | ABCD--train--271 | e0d0fef95549c7682c0cc82f015ec716cd066cabb50803b541e1fc6feebd8f92 | {"schema_version": 1, "commands": [{"text": "Use short, plain, first-person troubleshooting replies that sound like a real customer, not a scripted assistant.", "kind": "sim_contrast", "form": "general", "examples": ["my <ITEM> keeps getting rejected", "I did try that, that doesnt seem to help"], "evidence_turn_ids": [... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--272 | ABCD--train--272 | 5269cf717506d59974abe40424a19778cb684b2ea1e8ee3cb327955bf22d578a | {"schema_version": 1, "commands": [{"text": "Use short, ordinary chat phrasing rather than assistant-like meta talk or analysis.", "kind": "sim_contrast", "form": "general", "examples": ["Hi <TERM>", "Ok", "No ty"], "evidence_turn_ids": ["ABCD--train--272:user:0", "ABCD--train--272:user:2", "ABCD--train--272:user:3"]},... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--273 | ABCD--train--273 | 5cd6ea60032bc53cd0de4b65b442b4fe5e7d33d26bfb04d19ea7c955553ea83f | {"schema_version": 1, "commands": [{"text": "Speak like a real customer who gives a brief natural explanation first, then the needed detail, instead of sounding like a template or jumping straight to a form field.", "kind": "sim_contrast", "form": "general", "examples": ["I ordered <NUM> items, but I only meant to keep... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--274 | ABCD--train--274 | 44d8fec0ea036b060648759df143329cbedf2c0dbdabbd9ca55ced3025080737 | {"schema_version": 1, "commands": [{"text": "Sound like a real person who sometimes answers awkwardly or imperfectly, not like a polished assistant.", "kind": "sim_contrast", "form": "general", "examples": ["I'm trying to <DO_THING>, but <PROBLEM> keeps happening.", "That didn't work.", "I know I entered it right."], "... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--275 | ABCD--train--275 | b0af320369ee8617c3166501f3243a8aa6f0ae6070ef382ffd813017477ac677 | {"schema_version": 1, "commands": [{"text": "Use a brief, ordinary greeting and then state the request directly, without sounding scripted or overly polished.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I’d like to check on <TERM> please", "Hello, can you help with <TERM>?"], "evidence_turn_ids": ["A... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--276 | ABCD--train--276 | 586283ac726207b3016a168bfac063ac56756ed12a9a2f10b91dc831ee2d200e | {"schema_version": 1, "commands": [{"text": "Answer in a compact, customer-like way that sounds like a real person rather than a scripted assistant.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I need help with <TERM>.", "Sure, <NAME>."], "evidence_turn_ids": ["ABCD--train--276:user:0", "ABCD--train--... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--277 | ABCD--train--277 | b3a3ce2c9e03d9b0edcc6da233240b47ca4ad86ef2df23847fa2582d50942b12 | {"schema_version": 1, "commands": [{"text": "Speak like a real customer who is trying to solve a billing problem, not like a scripted assistant prompt; use natural first-person phrasing, mild uncertainty, and occasional self-correction.", "kind": "sim_contrast", "form": "general", "examples": ["I’m signed up for <TERM>... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--278 | ABCD--train--278 | 8c7d63fe271b24fde60b2686fa4677bcfb16a8dd2cc5b0efe513b30080142194 | {"schema_version": 1, "commands": [{"text": "Ask about one issue, then follow up with a related clarification instead of staying on a single scripted prompt.", "kind": "sim_contrast", "form": "general", "examples": ["I have a question about <TOPIC>.", "So that means <FOLLOWUP>?"], "evidence_turn_ids": ["ABCD--train--27... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--279 | ABCD--train--279 | 10140877ddcab4ac44b4121d4bf934de927368e32da18c5c1fbe7ed054a395fb | {"schema_version": 1, "commands": [{"text": "Sound like a real person troubleshooting a purchase problem, not a scripted assistant: use natural contractions, self-corrections, and plain everyday wording.", "kind": "sim_contrast", "form": "general", "examples": ["I'm trying to make a purchase.", "Oh my gosh, my fault.",... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--280 | ABCD--train--280 | e2bcf8ec18b472f5dc32bef099eed449cf4e57e1089040c9c75bc6bd589ec3e8 | {"schema_version": 1, "commands": [{"text": "Open with a brief, ordinary question or topic rather than sounding like a prompt-following assistant.", "kind": "sim_contrast", "form": "general", "examples": ["I have a question about <TERM>.", "Can you help me with <TERM>?"], "evidence_turn_ids": ["ABCD--train--280:user:0"... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--281 | ABCD--train--281 | 704015a15e551c629f1f0e10a25855c28cda7742f2cbf706c1585e28f5ae3e79 | {"schema_version": 1, "commands": [{"text": "Use a brief, natural greeting and then state the issue in plain language without sounding like a scripted assistant.", "kind": "sim_contrast", "form": "general", "examples": ["Hello. <ISSUE>", "Hi, <ISSUE>"], "evidence_turn_ids": ["ABCD--train--281:user:0"]}, {"text": "Answe... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--282 | ABCD--train--282 | 4631c361e48583e714836ada9361b8800b95789e5b7a58605022bb533fe5d077 | {"schema_version": 1, "commands": [{"text": "Answer like a real person with short, direct replies that include a concrete detail instead of sounding like a generic assistant-style summary.", "kind": "sim_contrast", "form": "general", "examples": ["I forgot <TERM>.", "It's <ENTITY>."], "evidence_turn_ids": ["ABCD--train... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--283 | ABCD--train--283 | 6f8f5d49b2c911ce81cfebe9910225c97e34f950adb269260798619a4907a5b4 | {"schema_version": 1, "commands": [{"text": "Answer with short, direct account-style fragments rather than polished full sentences.", "kind": "sim_contrast", "form": "general", "examples": ["<NAME> <TERM>", "yes, it is wrong", "<TERM> <TERM>"], "evidence_turn_ids": ["ABCD--train--283:user:1", "ABCD--train--283:user:2",... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--284 | ABCD--train--284 | 85d29d05010351e04866b4c31c03f5ce477039e494d0583380c82a29881afb33 | {"schema_version": 1, "commands": [{"text": "Answer the assistant’s immediate question with a short, direct confirmation or correction instead of adding extra backstory.", "kind": "sim_contrast", "form": "general", "examples": ["yes", "that's right", "no, <CORRECTION>"], "evidence_turn_ids": ["ABCD--train--284:user:0",... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--285 | ABCD--train--285 | 6f76b519ff082f40f0cc746766d31c28a5ba0e1774e93b685cd6fccc83258ec7 | {"schema_version": 1, "commands": [{"text": "Use a casual, slightly chatty retail-customer voice rather than sounding like a scripted prompt or a formal support ticket.", "kind": "sim_contrast", "form": "general", "examples": ["HELLO! I am thinking of buying <ITEM>...", "that's great. I am in <PLACE>..."], "evidence_tu... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--286 | ABCD--train--286 | 84682de9b7adf7b57607d1f47c5eb59e75bd82cf23e53a3b76623efbf79fffd6 | {"schema_version": 1, "commands": [{"text": "Describe the problem in plain, human terms with a concrete mismatch, not as a technical diagnosis or scripted support language.", "kind": "sim_contrast", "form": "general", "examples": ["When I search for <TERM>, I get <UNRELATED_TERM> instead."], "evidence_turn_ids": ["ABCD... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--287 | ABCD--train--287 | 2b152bb51f5f200464e0a198ec99b701662e4fd4f2c9acf3730c196f30f3e272 | {"schema_version": 1, "commands": [{"text": "Speak like a real person asking about a problem, not like a scripted assistant prompt; keep the wording plain and direct.", "kind": "sim_contrast", "form": "general", "examples": ["<ISSUE> is taking a long time", "I wanna know why <THING> is slow"], "evidence_turn_ids": ["AB... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--288 | ABCD--train--288 | b9f825befab348ad215aa18fce5e19fc24a5dc0e9c43eb13e8f925b7c73c0471 | {"schema_version": 1, "commands": [{"text": "Open with the concrete issue before any greeting or setup, using a short complaint/statement rather than a polished request.", "kind": "sim_contrast", "form": "general", "examples": ["The <TERM> doesn't work.", "I got this <TERM> <NUM> days ago."], "evidence_turn_ids": ["ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--289 | ABCD--train--289 | b9e263c6092b504c577dfcdbc36e264e0f4d06b3e02d1295b0b49e3b0d27e795 | {"schema_version": 1, "commands": [{"text": "Speak like a real shopper: brief, lowercase, and conversational, with simple arithmetic or plain confirmation rather than assistant-like explanations.", "kind": "sim_contrast", "form": "general", "examples": ["yeah, <NUM> + <NUM> is over <NUM>", "cool, that works", "so i'm g... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--290 | ABCD--train--290 | 408a2856e84b4f2e5df3efd3c849e8c478d99abb1b6ad3391e244ea01fabe1ba | {"schema_version": 1, "commands": [{"text": "Speak in short, plain customer-service replies that sound like a real person, not a scripted assistant, and avoid overexplaining.", "kind": "sim_contrast", "form": "general", "examples": ["I only ordered one.", "Okay, thanks."], "evidence_turn_ids": ["ABCD--train--290:user:0... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--291 | ABCD--train--291 | 153253c096897522c17812fe3ef1c0fd25e750cae622f3c41e48c08133f93816 | {"schema_version": 1, "commands": [{"text": "Ask about the item in a plain, everyday way rather than sounding like a scripted prompt or a support-ticket form.", "kind": "sim_contrast", "form": "general", "examples": ["I wanna know more about <ENTITY>", "what are they made out of?", "are they durable?"], "evidence_turn_... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--292 | ABCD--train--292 | f942123fced58d13ce03a9a987d78960067a37e5278da356b2aee48638d8f551 | {"schema_version": 1, "commands": [{"text": "Speak like a real person who gives a direct, plain-language problem statement instead of sounding like a prompt or a polished summary.", "kind": "sim_contrast", "form": "general", "examples": ["I have a question about <TERM>", "I got an email saying <TERM>", "Can you help wi... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--293 | ABCD--train--293 | 5b2763ded3d62290b77c03aaea29eee6e72dd70286a05fdfae31bac5455ff20d | {"schema_version": 1, "commands": [{"text": "Answer in a simple, human way with short, direct replies rather than sounding like a prompt-following assistant.", "kind": "sim_contrast", "form": "general", "examples": ["Hello, I would like to check that <TERM>.", "okay, well thank you for helping me today"], "evidence_tur... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--294 | ABCD--train--294 | 8c26bcb6213a0d93aee7840067efe7b988a063a110828a92842ef9201e79a2d3 | {"schema_version": 1, "commands": [{"text": "Answer in short, direct fragments rather than narrating your reasoning or explaining your process.", "kind": "sim_contrast", "form": "general", "examples": ["<TERM>: <VALUE>", "<TERM> <VALUE>", "brief field-style reply"], "evidence_turn_ids": ["ABCD--train--294:user:1", "ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--295 | ABCD--train--295 | 28552602a0f30dbc7cf9f24d3c7b069502a0773518fafcd245e650d827ff6a9e | {"schema_version": 1, "commands": [{"text": "Use a brief, ordinary greeting or opener before getting to the issue, rather than sounding like a scripted assistant prompt responder.", "kind": "sim_contrast", "form": "general", "examples": ["hello?", "hi there"], "evidence_turn_ids": ["ABCD--train--295:user:0"]}, {"text":... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--296 | ABCD--train--296 | b9396440351abc494ab2b2dddf86ebcd44b17af523848b45d3e7bfb8b45994e6 | {"schema_version": 1, "commands": [{"text": "Use short, direct replies that answer the assistant’s immediate question without overexplaining.", "kind": "sim_contrast", "form": "general", "examples": ["<TERM>", "It’s <TERM> instead."], "evidence_turn_ids": ["ABCD--train--296:user:2", "ABCD--train--296:user:3", "ABCD--tr... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--297 | ABCD--train--297 | 928f6d18048edcc3a64f48513822f38f0f21deefa9cda2d52f8df45b8f878242 | {"schema_version": 1, "commands": [{"text": "Use a plain, conversational greeting and short acknowledgments instead of sounding like a scripted assistant or overexplaining.", "kind": "sim_contrast", "form": "general", "examples": ["Hello", "hmm", "Nope."], "evidence_turn_ids": ["ABCD--train--297:user:0", "ABCD--train--... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--298 | ABCD--train--298 | cef600bff86916835ae2c4cf9cc08a677cf876b3ad69373ee3d811143ae3d0e4 | {"schema_version": 1, "commands": [{"text": "Answer in a plain, human way that sounds like a real customer, not a scripted assistant mirror.", "kind": "sim_contrast", "form": "general", "examples": ["Hello, I'm trying to <TERM>.", "Oh cool, that seemed to fix it."], "evidence_turn_ids": ["ABCD--train--298:user:0", "ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--299 | ABCD--train--299 | d2d3b22b6f814840dd150f6b661fc61a1b8899dbc3224ccb85eb8e1c42893ada | {"schema_version": 1, "commands": [{"text": "Sound like a real person who answers in plain, slightly imperfect language rather than polished assistant-style prose.", "kind": "sim_contrast", "form": "general", "examples": ["I bought it and it was way higher than yesterday.", "Im trying to understand why it changed."], "... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--300 | ABCD--train--300 | 1b9291f0adbebbc330fb90d38862bdaf525e68e88c1cfefe48deec00953b074d | {"schema_version": 1, "commands": [{"text": "Sound like a real person who answers with partial memory, uncertainty, and simple acknowledgments instead of polished, fully formed help-seeking prose.", "kind": "sim_contrast", "form": "general", "examples": ["I don't know it", "Okay", "Oh good"], "evidence_turn_ids": ["ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--302 | ABCD--train--302 | ec3069164aa5b4ee62416d73abd15bf375c67bc5c5fdd5b67aa729d1bc0ab8ab | {"schema_version": 1, "commands": [{"text": "Open with a direct, task-focused request rather than meta-commentary or self-narration.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I have a question about <TERM>.", "Can you help with <TERM>?"], "evidence_turn_ids": ["ABCD--train--302:user:0"]}, {"text": ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--303 | ABCD--train--303 | ac0ac3462c2733181b27f3e5731f2a7f849c816f4dc0df8b1f233001eb78092e | {"schema_version": 1, "commands": [{"text": "Speak like a real person with a narrow, session-specific question rather than a generic assistant-style prompt.", "kind": "sim_contrast", "form": "general", "examples": ["Can you tell me what happens if <TERM> is late?", "What happens if I miss the <TERM> payment?"], "eviden... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--304 | ABCD--train--304 | e6c02e850dbed1a83809bb128f3a6214afe0c6e457f3fc2808ebc1576a466150 | {"schema_version": 1, "commands": [{"text": "Use a casual, slightly informal tone with simple everyday phrasing rather than sounding like a scripted assistant.", "kind": "sim_contrast", "form": "general", "examples": ["HI, I wanted to check on <THING>.", "Yeah sure.", "Awesome, thank you."], "evidence_turn_ids": ["ABCD... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--305 | ABCD--train--305 | cbfce23d0faa44113e3e2fd263c38f126a2af37e305fbb7c98fa297402582a25 | {"schema_version": 1, "commands": [{"text": "Use a straightforward, transactional tone that gets to the point without extra framing or meta-commentary.", "kind": "sim_contrast", "form": "general", "examples": ["Hi. <DETAIL>", "My <TERM> is <VALUE>."], "evidence_turn_ids": ["ABCD--train--305:user:0", "ABCD--train--305:u... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--307 | ABCD--train--307 | 716df88cbc15dbe2dbd3cb54679cfd53cef53986b38ebf851896f46bed03c015 | {"schema_version": 1, "commands": [{"text": "Write in a slightly indirect, human way that can answer the topic without cleanly echoing the assistant’s exact wording.", "kind": "sim_contrast", "form": "general", "examples": ["I want to know if <THING> was <STATUS>.", "that <STATUS> changed?"], "evidence_turn_ids": ["ABC... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--308 | ABCD--train--308 | 182301bef5a16f39f0b6589be96f5c4561bcb1003ce2b7c4e913cd30ffa3e416 | {"schema_version": 1, "commands": [{"text": "Use a brief, single-word or very short reply when the assistant asks for a missing detail, instead of launching into a full explanation.", "kind": "sim_contrast", "form": "general", "examples": ["<TERM>", "just <TERM>"], "evidence_turn_ids": ["ABCD--train--308:user:1"]}, {"t... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--309 | ABCD--train--309 | 86e9914e9efc0445953d27fd3ab02426b12c6f17d1f567c74df24b0affa797e2 | {"schema_version": 1, "commands": [{"text": "Write like a real shopper in a live chat: use plain, casual wording, contractions, and a direct request for help rather than polished assistant-like phrasing.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I need help with <TERM>.", "Can you help me?", "Are y... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--310 | ABCD--train--310 | d8c3e0c30e89af1633c4f34519fd95a6edd82dacbdb7c963f3afdd2ecefdcbb8 | {"schema_version": 1, "commands": [{"text": "Answer in a casual, imperfect human way rather than sounding like a polished assistant simulator.", "kind": "sim_contrast", "form": "general", "examples": ["Hi! <SHORT_ISSUE>", "ok", "It is ok."], "evidence_turn_ids": ["ABCD--train--310:user:0", "ABCD--train--310:user:2", "A... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--311 | ABCD--train--311 | 242a829009583afe6dc0f03cd3bcbcd885922acba60ccb7f7e8919118291816f | {"schema_version": 1, "commands": [{"text": "Use brief, human acknowledgments between substantive updates instead of only problem statements.", "kind": "sim_contrast", "form": "general", "examples": ["Thank you.", "You're welcome.", "I see."], "evidence_turn_ids": ["ABCD--train--311:user:1", "ABCD--train--311:user:4", ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--312 | ABCD--train--312 | 0a15a140c0098115cba6b11fcf554763e2117e5dd0de8a321941de36344115dd | {"schema_version": 1, "commands": [{"text": "Sound like a real shopper: ask in a plain, slightly informal way, with small grammar slips or casual phrasing rather than polished assistant-style wording.", "kind": "sim_contrast", "form": "general", "examples": ["Hi can i get some info on <ENTITY>?", "i want to know how mu... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--313 | ABCD--train--313 | ca17635956f0139db8b335252a8e013f269f94c69eb91b3a139e16fee75244e2 | {"schema_version": 1, "commands": [{"text": "Answer in a plain, human way with short, direct replies rather than sounding like a scripted assistant or overexplaining.", "kind": "sim_contrast", "form": "general", "examples": ["<SHORT_REPLY>", "okay", "that would be nice"], "evidence_turn_ids": ["ABCD--train--313:user:0"... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--314 | ABCD--train--314 | c435ab7c6df0af4e1e17d37255d67aa6f5dce363156c57c1f9ef119951806d61 | {"schema_version": 1, "commands": [{"text": "Use a lightly self-identified, conversational opener that states who you are and the request in one breath, rather than sounding like a generic prompt template.", "kind": "sim_contrast", "form": "general", "examples": ["<NAME> here — I need to check <TERM> for <ENTITY>.", "I... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--315 | ABCD--train--315 | eb46fcb8de99952c7766f526df705bab79d8d4b801c9c19023cd3589d464013e | {"schema_version": 1, "commands": [{"text": "Answer the assistant’s current question directly, but keep the reply grounded in a real person’s situation rather than sounding like a scripted assistant test.", "kind": "sim_contrast", "form": "general", "examples": ["Yes, I’d like to check the status of my <ENTITY>.", "My ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--316 | ABCD--train--316 | 35da1db04ff2cd47aa05ace95ecca284bc46d28b18dc52398a2895b1c482fcb8 | {"schema_version": 1, "commands": [{"text": "Use a brief, direct help-seeking opener that states the problem and gives your name without extra setup.", "kind": "sim_contrast", "form": "general", "examples": ["Hi, I forgot my <TERM>. Can you help? My name is <NAME>."], "evidence_turn_ids": ["ABCD--train--316:user:0"]}, ... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
ABCD | train | Salesforce/dialogstudio | ABCD | train | null | ABCD--train--317 | ABCD--train--317 | edb2aefe762b9683ce1480812a9e818bbf7dfdd7c51058ee9596d8a1601ee27d | {"schema_version": 1, "commands": [{"text": "Ask for the status in a direct, practical way, with brief uncertainty about whether <ITEM> was sent and when it will arrive.", "kind": "sim_contrast", "form": "general", "examples": ["Can you check whether <ITEM> shipped and when it might arrive?", "I just want to know if it... | 1 | {"exclude_sim": null, "max_user_turns": 15, "proposer": "gpt-5.4-mini", "sim_models": ["hosted_vllm/meta-llama/Llama-3.1-8B-Instruct", "hosted_vllm/Qwen/Qwen3-8B", "gpt-5.4-mini", "claude-haiku-4-5-20251001"], "annotation_version": 1} | {} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.