WebJev 1.0

.gitattributes

@@ -0,0 +1,60 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.avro filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mds filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
+# Video files - compressed
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.webm filter=lfs diff=lfs merge=lfs -text

LICENSE

@@ -0,0 +1,11 @@
+WebJev
+Copyright (c) 2026 Lexmount. All rights reserved.
+
+This dataset is proprietary and confidential. It is made available only to members of the Lexmount organization on
+Hugging Face, for Lexmount's internal research and model development. Do not redistribute, publish or share the
+dataset, or any substantial part of it, outside Lexmount without written permission from Lexmount.
+
+Third-party content. Page text, URLs and element labels were captured from publicly accessible websites and remain
+the property of their respective owners. Task descriptions derive from third-party task collections that carry their
+own licenses. Labels were produced with DeepSeek models under their terms of use. Nothing in this repository grants
+any right in third-party content.

README.md

@@ -0,0 +1,448 @@
+---
+pretty_name: WebJev
+license: other
+license_name: lexmount-internal
+license_link: LICENSE
+language:
+- en
+multilinguality: monolingual
+size_categories:
+- 10K<n<100K
+source_datasets:
+- original
+annotations_creators:
+- machine-generated
+language_creators:
+- found
+- machine-generated
+task_categories:
+- multiple-choice
+- text-classification
+task_ids:
+- multiple-choice-qa
+tags:
+- web-agent
+- browser-agent
+- gui-agent
+- action-prediction
+- element-grounding
+- decision-making
+- live-web
+configs:
+- config_name: all
+  default: true
+  data_files:
+  - split: train
+    path: data/*/train-*.parquet
+- config_name: action_prediction
+  data_files:
+  - split: train
+    path: data/action_prediction/train-*.parquet
+- config_name: element_grounding
+  data_files:
+  - split: train
+    path: data/element_grounding/train-*.parquet
+dataset_info:
+- config_name: all
+  features: &id001
+  - name: id
+    dtype: string
+  - name: task_type
+    dtype: string
+  - name: question_type
+    dtype: string
+  - name: action
+    dtype: string
+  - name: context
+    dtype: string
+  - name: question
+    dtype: string
+  - name: options
+    list: string
+  - name: gold
+    dtype: int32
+  - name: answer
+    dtype: string
+  - name: num_options
+    dtype: int32
+  - name: num_tokens
+    dtype: int32
+  - name: context_format
+    dtype: string
+  - name: website
+    dtype: string
+  - name: page_url
+    dtype: string
+  - name: task_source
+    dtype: string
+  - name: label_source
+    dtype: string
+  - name: task_id
+    dtype: string
+  - name: instance_id
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 1139312203
+    num_examples: 64122
+  download_size: 120376472
+  dataset_size: 1139312203
+- config_name: action_prediction
+  features: *id001
+  splits:
+  - name: train
+    num_bytes: 395573253
+    num_examples: 31050
+  download_size: 33971063
+  dataset_size: 395573253
+- config_name: element_grounding
+  features: *id001
+  splits:
+  - name: train
+    num_bytes: 743738950
+    num_examples: 33072
+  download_size: 86405409
+  dataset_size: 743738950
+---
+
+# WebJev
+
+**Training web agents to complete real tasks on the live web**
+
+To complete a task on a real website, a web agent must get a long chain of decisions right, from the first page to
+the final answer: *what to do next*, and *which element to act on*. WebJev captures these decisions along complete
+task trajectories on the live web. It contains **64,122 decisions** from **3,858 tasks** on
+**1,443 real-world websites**. They cover every stage of a task: searching, navigating, filtering, filling in
+forms, reading results and deciding when the task is done.
+
+Each instance pairs an agent's observation of the current page with one multiple-choice question. Together the
+instances cover two complementary skills:
+
+- **Action prediction** (31,050 instances): choose the next operation among those valid on the current page
+  (3–13 options). The operations are click, type, select, scroll, wait, submit, dismiss, go back, finish and give up.
+- **Element grounding** (33,072 instances): choose the element to act on among up to 137
+  candidates from the same page (median 32).
+
+Observations follow the format a deployed browser agent sends to its decision model. They include:
+
+- the page's URL, title and visible text;
+- the actionable elements in the viewport, with their roles, values and positions;
+- the recent action history.
+
+An LLM annotator produced the labels while operating the agent on live websites, and only labels that passed
+multi-signal verification were kept. The grounding instances are further enriched with same-page hard negatives at
+several candidate scales.
+
+![One task, completed decision by decision](assets/trajectory.png)
+*One WebJev task on spanishdict.com, completed decision by decision: type the query, pick the suggestion, dismiss a
+pop-up, read the entry and finish. The blue boxes mark the elements the agent acts on.*
+
+## Highlights
+
+- **Whole-task coverage.** Decisions come from complete task trajectories on live websites. They span every stage of
+  a task, from the first search to the final `DONE`.
+- **Live and diverse.** 3,858 tasks on 1,443 real-world websites, from shopping, travel and real estate
+  to education, research, software and sports.
+- **Two decision skills.** 31,050 action-prediction and 33,072 element-grounding instances.
+- **Deployment-faithful observations.** 89% of the instances use the structured observation of a
+  production browser agent; the rest use a textual page rendering that adds format diversity.
+- **Hard grounding.** Grounding questions have a median of 32 candidates and up to
+  137, with same-page hard negatives and multi-scale candidate sets.
+- **Long context.** The median instance has 4,262 tokens, and 10% of instances have more than
+  9,730.
+- **Verified labels.** Labels are kept only after multi-signal verification: execution on the live page,
+  re-annotation consistency and adjudication. Independent audits of stratified samples found 98% of annotated labels
+  correct or acceptable.
+
+| At a glance | |
+|---|---|
+| Instances | 64,122 (31,050 action prediction, 33,072 element grounding) |
+| Tasks · websites | 3,858 · 1,443 |
+| Distinct decision points | 40,230 |
+| Candidates per question | action: 3–13 (median 7); grounding: 2–137 (median 32) |
+| Context length | median 4,262 tokens, 90th percentile 9,730, maximum 16,383 (327.6M in total) |
+| Observation formats | structured agent observation 57,283; textual rendering 6,839 |
+| Language | English tasks and instructions; page text mostly English |
+| Collection period | August–September 2026 |
+
+![Grounding examples from twelve websites](assets/gallery.png)
+*Element-grounding examples from twelve of the 1,443 websites. The blue box marks the gold element; each caption
+gives the intended step and the number of candidates.*
+
+## Task formulation
+
+A web task unfolds as a chain of decisions. At every step the agent observes the page, decides what to do and, for
+clicks, typing and selects, which element to act on. Each decision yields one instance per question.
+
+![Anatomy of one decision](assets/overview.png)
+*One decision on spanishdict.com. After typing "spring" into the search box, the agent must decide what to do next
+(action prediction) and which of the 28 visible elements to click (element grounding).*
+
+Each instance is a tuple *(observation, instruction, candidates, answer)*:
+
+- **Observation** (`context`). What the agent sees at this step. The *structured* form is a JSON object:
+  - `page`: URL, title and visible text;
+  - `elements`: actionable elements of the viewport, each with its index, role, label, current value, supported
+    operations and position;
+  - `recent_actions`: the recent action history and whether each action changed the page.
+
+  The *textual* form renders the website, the task, the current subgoal, the actions taken so far and the page text
+  as plain text.
+- **Instruction** (`question`). What is asked. Structured instances carry the user's task, the current subgoal, the
+  date and the agent's decision policy. Textual instances carry a single question.
+- **Candidates** (`options`). For action prediction, the operations valid on the page. For element grounding, the
+  elements of the page that support the requested operation.
+- **Answer** (`gold`). The index of the correct candidate.
+
+**Action space.**
+
+| Action | Meaning |
+|---|---|
+| `CLICK` | click a link, button, menu item, suggestion or date |
+| `TYPE_TEXT` | enter text into an editable field |
+| `SELECT` | choose a value in a drop-down |
+| `SCROLL_DOWN` / `SCROLL_UP` | scroll the page |
+| `SCROLL_REGION` | scroll inside a scrollable page region (appears as `E<n>` in the options) |
+| `WAIT` | wait for the page to update |
+| `PRESS_ENTER` | submit the focused field |
+| `PRESS_ESCAPE` | close a menu or pop-up |
+| `GO_BACK` | return to the previous page |
+| `DONE` | the goal is visibly satisfied |
+| `BLOCKED` | no available action can make progress |
+
+**Model interface.** The dataset suits decision models that score options rather than generate free text. We render
+each instance as follows:
+
+```text
+Context:
+{context}
+
+Question: {question}
+Options:
+(A) {option}
+(B) {option}
+…
+Answer: (
+```
+
+- The option order is shuffled.
+- Options are labelled `A`, `B`, … and then `AA`, `AB`, … for large candidate sets.
+- The model is trained with cross-entropy over the option labels at the answer position.
+- `num_tokens` reports the length of this rendering under the Qwen3.5 tokenizer.
+
+## Data distribution
+
+![Distributions of WebJev](assets/distributions.png)
+*(a) Action labels of action-prediction instances. (b) Composition by question type. (c) Candidate-set size of
+element-grounding instances. (d) Context length.*
+
+### Composition
+
+| Task type | Question type | Structured observation | Textual observation | Total |
+|---|---|---:|---:|---:|
+| Action prediction | `operation` | 31,050 | 0 | 31,050 |
+| Element grounding | `click_target` | 25,465 | 6,197 | 31,662 |
+| Element grounding | `type_target` | 541 | 604 | 1,145 |
+| Element grounding | `select_target` | 227 | 38 | 265 |
+| **Total** | | **57,283** | **6,839** | **64,122** |
+
+### Actions
+
+The action labels follow the natural frequency of decisions during web navigation. Clicks and scrolling dominate, and
+terminal decisions (`DONE`, `BLOCKED`) account for 8.6%.
+
+| Action | Instances | Share |
+|---|---:|---:|
+| `CLICK` | 14,364 | 46.3% |
+| `SCROLL_DOWN` | 6,135 | 19.8% |
+| `TYPE_TEXT` | 3,527 | 11.4% |
+| `DONE` | 2,397 | 7.7% |
+| `GO_BACK` | 1,553 | 5.0% |
+| `WAIT` | 1,137 | 3.7% |
+| `SCROLL_UP` | 713 | 2.3% |
+| `PRESS_ENTER` | 362 | 1.2% |
+| `SCROLL_REGION` | 355 | 1.1% |
+| `BLOCKED` | 259 | 0.8% |
+| `SELECT` | 226 | 0.7% |
+| `PRESS_ESCAPE` | 22 | 0.1% |
+
+### Grounding difficulty
+
+Grounding questions are hard: the correct element must be chosen among many look-alike candidates of the same page.
+
+- **Click targets** are the largest group. Their candidate sets are multi-scale (32 / 64 / 128 / all elements), which
+  exposes models to increasing distractor density.
+- **Select targets** often list long option menus.
+- **Type targets** usually choose between a few input fields.
+
+| Question type | Instances | Min | Median | 90th percentile | Max |
+|---|---:|---:|---:|---:|---:|
+| `operation` (action prediction) | 31,050 | 3 | 7 | 9 | 13 |
+| `click_target` | 31,662 | 2 | 32 | 86 | 137 |
+| `type_target` | 1,145 | 2 | 2 | 4 | 12 |
+| `select_target` | 265 | 2 | 26 | 81 | 131 |
+| all grounding | 33,072 | 2 | 32 | 84 | 137 |
+
+### Context length
+
+| Subset | Instances | Median | 90th percentile | Max |
+|---|---:|---:|---:|---:|
+| Action prediction | 31,050 | 3,056 | 6,902 | 16,371 |
+| Element grounding | 33,072 | 5,665 | 11,366 | 16,383 |
+| Structured observations | 57,283 | 4,102 | 9,332 | 16,383 |
+| Textual observations | 6,839 | 6,102 | 11,587 | 16,376 |
+| **All** | 64,122 | 4,262 | 9,730 | 16,383 |
+
+### Language
+
+- The tasks, instructions and candidates are in English.
+- Page text is predominantly English. In 0.9% of the instances a non-Latin script makes up at least 20%
+  of the observation: CJK 434, Cyrillic 65, Devanagari 53, Arabic 36, Greek 5.
+- Localized pages in other Latin-script languages also occur.
+
+### Label sources
+
+| Label source | Instances | Share | Description |
+|---|---:|---:|---|
+| `live_annotation` | 43,890 | 68.4% | action chosen by the LLM annotator on the live page, kept after verification |
+| `trajectory` | 20,011 | 31.2% | target taken from a successful agent trajectory and verified |
+| `trajectory_corrected` | 221 | 0.3% | trajectory target replaced by the reviewer's corrected target |
+
+## Dataset structure
+
+### Configs and splits
+
+| Config | Instances | Content |
+|---|---:|---|
+| `all` (default) | 64,122 | all instances |
+| `action_prediction` | 31,050 | next-action questions |
+| `element_grounding` | 33,072 | click, type and select target questions |
+
+The dataset is released as a single `train` split. Instances that share a `task_id` come from the same task, and
+instances that share an `instance_id` come from the same decision step: its action and grounding questions, or the
+same grounding question at different candidate scales. Use `task_id` to build leakage-free held-out splits.
+
+### Data fields
+
+| Field | Type | Description |
+|---|---|---|
+| `id` | string | unique instance id |
+| `task_type` | string | `action_prediction` or `element_grounding` |
+| `question_type` | string | `operation`, `click_target`, `type_target` or `select_target` |
+| `action` | string | action prediction: the gold action (see the action space). Grounding: the operation being grounded |
+| `context` | string | the observation (model input) |
+| `question` | string | the instruction and question (model input) |
+| `options` | list[string] | the candidates (model input) |
+| `gold` | int32 | index of the correct candidate in `options` |
+| `answer` | string | `options[gold]` |
+| `num_options` | int32 | number of candidates |
+| `num_tokens` | int32 | length of the rendered instance (Qwen3.5 tokenizer) |
+| `context_format` | string | `structured` or `text` |
+| `website` | string | the task's target website |
+| `page_url` | string, nullable | URL of the observed page (structured observations) |
+| `task_source` | string | public task collection of the task |
+| `label_source` | string | `live_annotation`, `trajectory` or `trajectory_corrected` (see Label sources) |
+| `task_id` | string | task identifier, for grouping |
+| `instance_id` | string | decision-step identifier, for grouping |
+
+Only `context`, `question` and `options` are model inputs, and `gold` is the target. The remaining fields describe
+the instance and are meant for analysis and splitting.
+
+### Example
+
+```json
+{
+  "task_type": "element_grounding",
+  "question_type": "click_target",
+  "action": "CLICK",
+  "context": "{\"page\": {\"url\": \"https://www.spanishdict.com/\", \"title\": \"SpanishDictionary.com | English to Spanish Translation, Dictionary, Translator\", \"text\": \"Learn Spanish\\nTranslation\\nConjugation\\n…\"}, \"elements\": [{\"role\": \"link\", \"index\": \"1\", \"label\": \"SpanishDictionary.com Homepage\", \"operations\": [\"CLICK\"], \"position\": {\"x\": 213, \"y\": 0}}, …], \"recent_actions\": [{\"action\": \"Translate Spanish or English\", \"kind\": \"fill\", \"text\": \"spring\", \"page_changed\": true}]}",
+  "question": "{\"goal\": \"… The user's original task: On https://www.spanishdict.com: Search for the word 'spring' on SpanishDict and identify two different Spanish translations that represent distinct meanings. …\", \"operation\": \"CLICK\", \"rules\": [\"Advance the user's entire goal from the CURRENT page using one operation. …\"]}",
+  "options": ["1: {\"element\": \"[1] SpanishDictionary.com Homepage\", …}", "…", "14: {\"element\": \"[14] spring\", \"role\": \"option\", …}", "15: {\"element\": \"[15] spring break\", \"role\": \"option\", …}", "…"],
+  "gold": 12,
+  "answer": "14: {\"element\": \"[14] spring\", \"current_value\": \"0\", \"role\": \"option\", \"selected\": \"false\"}",
+  "num_options": 28,
+  "num_tokens": 3949,
+  "context_format": "structured",
+  "website": "spanishdict.com",
+  "task_source": "MolmoWeb",
+  "label_source": "live_annotation"
+}
+```
+
+### Usage
+
+```python
+from datasets import load_dataset
+
+ds = load_dataset("Lexmount/WebJev", "all", split="train")    # private: log in with a Lexmount account
+grounding = load_dataset("Lexmount/WebJev", "element_grounding", split="train")
+
+def render(ex):
+    labels = [chr(65 + i) for i in range(26)] + [a + b for a in map(chr, range(65, 91)) for b in map(chr, range(65, 91))]
+    options = "".join(f"\n({labels[i]}) {o}" for i, o in enumerate(ex["options"]))
+    return f"Context:\n{ex['context']}\n\nQuestion: {ex['question']}\nOptions:{options}\nAnswer: ("
+
+# leakage-free split by task
+tasks = sorted(set(ds["task_id"]))
+held_out = set(tasks[: len(tasks) // 20])
+train = ds.filter(lambda ex: ex["task_id"] not in held_out)
+valid = ds.filter(lambda ex: ex["task_id"] in held_out)
+```
+
+## Dataset construction
+
+![Construction of WebJev](assets/pipeline.png)
+
+WebJev was built in four stages.
+
+1. **Task curation.** We build on the task collections of MolmoWeb and WebGym. We keep tasks on real, publicly
+   reachable websites that need no account.
+2. **Agent-in-the-loop collection.** An LLM annotator (DeepSeek-V4.1-Flash) operated a production-grade browser
+   agent on the live websites, in isolated cloud browser sessions.
+   - At every step, the agent's observation and questions were recorded exactly as served.
+   - The annotator's choice was then executed, and its effect on the page was recorded with it.
+   - A successful demonstration of the same task guided the annotator. It is never part of the model input, so each
+     instance must be answerable from the observation alone.
+3. **Multi-signal verification.** Each candidate label was checked against complementary signals:
+   - execution evidence and a meaningful page-state change;
+   - trajectory-level loop detection;
+   - agreement under independent re-annotation with shuffled options and without guidance;
+   - adjudication by an independent verifier for terminal and ambiguous decisions.
+
+   A final review restored valid decisions that conservative checks had rejected and removed near-duplicates. About
+   half of all recorded decisions survived verification.
+4. **Grounding enrichment.** Successful steps of agent trajectories on the same websites were converted into
+   additional element-grounding instances.
+   - Candidate sets are built from the actionable elements of the same page, with mined hard negatives.
+   - Each question is rendered at several candidate scales.
+   - An LLM reviewer checks the target, and a blind challenge removes candidates that could also be correct.
+   - A textual observation view adds format diversity.
+
+**Quality control.**
+
+- Independent audits of stratified samples, checked against page screenshots, found 98% of the annotated labels
+  correct or acceptable.
+- Every instance fits in 16K tokens without truncation.
+- Exact duplicates were removed.
+
+## Licensing and access
+
+- **Licence.** WebJev is proprietary to Lexmount and shared privately for internal research and development;
+  see [LICENSE](LICENSE).
+- **Third-party rights.** Page content belongs to the respective websites. The task descriptions derive from public
+  task collections under their own licences.
+
+## Citation
+
+```bibtex
+@misc{lexmount2026webjev,
+  title        = {WebJev: Training Web Agents to Complete Real Tasks on the Live Web},
+  author       = {{Lexmount}},
+  year         = {2026},
+  howpublished = {\url{https://huggingface.co/datasets/Lexmount/WebJev}}
+}
+```
+
+## Contact
+
+Lexmount, via the Lexmount organization on Hugging Face.

SHA256SUMS

@@ -0,0 +1,16 @@
+9e635ccfe0655b5809f89b3ede14bdc43220c3c8ceb9ca31815155e33c4fd99b  LICENSE
+2156e2e6b31314b4b17c517c0410154d4f6ad787cdc21a4b57105e7dbb2863d3  README.md
+5451fb4d3263d5d285cd8628f469a54656a6fe76560b0761ece95c8b09a931a8  assets/distributions.png
+be04fb8f321213fea00796d3501819197351b3ced6ce35e9f27d0ae3a4fdabec  assets/gallery.png
+5909605b8de1cec98ccd8ad666e43335c27420627e222597a49c07809950de78  assets/overview.png
+f89c5d8501daba1fa52a365acf12ebfa792bbaa94d7ddc24bdd35194d8341da8  assets/pipeline.png
+687a25a8fc4499c3fd1c689c879cae548b5e3b844bcb573ec12f7b07217b2757  assets/trajectory.png
+f68753a046fcda038a499ba0136f8605d59a38d63d7c512d378585074e3efaa2  data/action_prediction/train-00000-of-00004.parquet
+da9b7c7f190ba0663dc419d171fdb4d0c69a63e3309b8739769c5dc9909b74e3  data/action_prediction/train-00001-of-00004.parquet
+3a9bc280eab424e92170c4eb49c613801c06c55731a54c04b4c9c56d385e3e5a  data/action_prediction/train-00002-of-00004.parquet
+f3b8bdbd451959b4442cf0200cb699c8cf93bf5578ce77d0376834dd75f1c75c  data/action_prediction/train-00003-of-00004.parquet
+e59b893d5d4514a820c26750a087ad611367018c70601a71e567ee8792d9d307  data/element_grounding/train-00000-of-00004.parquet
+09ef97147ac908209c841a1cebb3baf0aa14944f811c6dc478fef0a2af1f661a  data/element_grounding/train-00001-of-00004.parquet
+baf3c6ac063fc99f275aa5f8ab0f242e580df243513e9ef615024cc783b1e46d  data/element_grounding/train-00002-of-00004.parquet
+885e088fb583b6229ba3b12f2341d683c60eeba44bf350b155c6eda9f17564a8  data/element_grounding/train-00003-of-00004.parquet
+5c8dba0e46feef7d7852dd1bb0e7b82c897ab2eec7dfbcd1c3061573960ecbbd  statistics.json

data/action_prediction/train-00000-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f68753a046fcda038a499ba0136f8605d59a38d63d7c512d378585074e3efaa2
+size 10887647

data/action_prediction/train-00001-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da9b7c7f190ba0663dc419d171fdb4d0c69a63e3309b8739769c5dc9909b74e3
+size 10958376

data/action_prediction/train-00002-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a9bc280eab424e92170c4eb49c613801c06c55731a54c04b4c9c56d385e3e5a
+size 10950629

data/action_prediction/train-00003-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3b8bdbd451959b4442cf0200cb699c8cf93bf5578ce77d0376834dd75f1c75c
+size 1174411

data/element_grounding/train-00000-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e59b893d5d4514a820c26750a087ad611367018c70601a71e567ee8792d9d307
+size 26236249

data/element_grounding/train-00001-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09ef97147ac908209c841a1cebb3baf0aa14944f811c6dc478fef0a2af1f661a
+size 25995570

data/element_grounding/train-00002-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baf3c6ac063fc99f275aa5f8ab0f242e580df243513e9ef615024cc783b1e46d
+size 26111391

data/element_grounding/train-00003-of-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:885e088fb583b6229ba3b12f2341d683c60eeba44bf350b155c6eda9f17564a8
+size 8062199

statistics.json

@@ -0,0 +1,316 @@
+{
+ "rows": 64122,
+ "by_task_type": {
+  "element_grounding": 33072,
+  "action_prediction": 31050
+ },
+ "by_question_type": {
+  "click_target": 31662,
+  "operation": 31050,
+  "type_target": 1145,
+  "select_target": 265
+ },
+ "action_prediction_labels": {
+  "CLICK": 14364,
+  "SCROLL_DOWN": 6135,
+  "TYPE_TEXT": 3527,
+  "DONE": 2397,
+  "GO_BACK": 1553,
+  "WAIT": 1137,
+  "SCROLL_UP": 713,
+  "PRESS_ENTER": 362,
+  "SCROLL_REGION": 355,
+  "BLOCKED": 259,
+  "SELECT": 226,
+  "PRESS_ESCAPE": 22
+ },
+ "element_grounding_actions": {
+  "CLICK": 31662,
+  "TYPE_TEXT": 1145,
+  "SELECT": 265
+ },
+ "by_context_format": {
+  "structured": 57283,
+  "text": 6839
+ },
+ "by_context_format_and_task_type": {
+  "structured|action_prediction": 31050,
+  "structured|element_grounding": 26233,
+  "text|element_grounding": 6839
+ },
+ "by_question_type_and_format": {
+  "operation|structured": 31050,
+  "click_target|structured": 25465,
+  "click_target|text": 6197,
+  "type_target|text": 604,
+  "type_target|structured": 541,
+  "select_target|structured": 227,
+  "select_target|text": 38
+ },
+ "by_label_source": {
+  "live_annotation": 43890,
+  "trajectory": 20011,
+  "trajectory_corrected": 221
+ },
+ "by_task_source": {
+  "MolmoWeb": 47845,
+  "WebGym": 16277
+ },
+ "by_page_script": {
+  "Latin": 63529,
+  "CJK": 434,
+  "Cyrillic": 65,
+  "Devanagari": 53,
+  "Arabic": 36,
+  "Greek": 5
+ },
+ "num_options": {
+  "all": {
+   "n": 64122,
+   "min": 2,
+   "p25": 7,
+   "p50": 9,
+   "p75": 32,
+   "p90": 64,
+   "p99": 128,
+   "max": 137,
+   "mean": 26.8,
+   "sum": 1719824
+  },
+  "element_grounding": {
+   "n": 33072,
+   "min": 2,
+   "p25": 27,
+   "p50": 32,
+   "p75": 64,
+   "p90": 84,
+   "p99": 128,
+   "max": 137,
+   "mean": 45.6,
+   "sum": 1506705
+  },
+  "click_target": {
+   "n": 31662,
+   "min": 2,
+   "p25": 31,
+   "p50": 32,
+   "p75": 64,
+   "p90": 86,
+   "p99": 128,
+   "max": 137,
+   "mean": 47.2,
+   "sum": 1494326
+  },
+  "structured": {
+   "n": 57283,
+   "min": 2,
+   "p25": 7,
+   "p50": 8,
+   "p75": 32,
+   "p90": 64,
+   "p99": 128,
+   "max": 137,
+   "mean": 23.0,
+   "sum": 1319732
+  },
+  "action_prediction": {
+   "n": 31050,
+   "min": 3,
+   "p25": 6,
+   "p50": 7,
+   "p75": 8,
+   "p90": 9,
+   "p99": 10,
+   "max": 13,
+   "mean": 6.9,
+   "sum": 213119
+  },
+  "operation": {
+   "n": 31050,
+   "min": 3,
+   "p25": 6,
+   "p50": 7,
+   "p75": 8,
+   "p90": 9,
+   "p99": 10,
+   "max": 13,
+   "mean": 6.9,
+   "sum": 213119
+  },
+  "text": {
+   "n": 6839,
+   "min": 2,
+   "p25": 32,
+   "p50": 64,
+   "p75": 64,
+   "p90": 128,
+   "p99": 128,
+   "max": 128,
+   "mean": 58.5,
+   "sum": 400092
+  },
+  "type_target": {
+   "n": 1145,
+   "min": 2,
+   "p25": 2,
+   "p50": 2,
+   "p75": 3,
+   "p90": 4,
+   "p99": 11,
+   "max": 12,
+   "mean": 2.8,
+   "sum": 3238
+  },
+  "select_target": {
+   "n": 265,
+   "min": 2,
+   "p25": 7,
+   "p50": 26,
+   "p75": 54,
+   "p90": 81,
+   "p99": 123,
+   "max": 131,
+   "mean": 34.5,
+   "sum": 9141
+  }
+ },
+ "num_tokens": {
+  "all": {
+   "n": 64122,
+   "min": 77,
+   "p25": 2662,
+   "p50": 4262,
+   "p75": 6852,
+   "p90": 9730,
+   "p99": 15328,
+   "max": 16383,
+   "mean": 5108.6,
+   "sum": 327573009
+  },
+  "element_grounding": {
+   "n": 33072,
+   "min": 77,
+   "p25": 3917,
+   "p50": 5665,
+   "p75": 8444,
+   "p90": 11366,
+   "p99": 15726,
+   "max": 16383,
+   "mean": 6418.8,
+   "sum": 212283887
+  },
+  "click_target": {
+   "n": 31662,
+   "min": 315,
+   "p25": 4028,
+   "p50": 5791,
+   "p75": 8538,
+   "p90": 11463,
+   "p99": 15738,
+   "max": 16383,
+   "mean": 6564.8,
+   "sum": 207853294
+  },
+  "structured": {
+   "n": 57283,
+   "min": 891,
+   "p25": 2595,
+   "p50": 4102,
+   "p75": 6500,
+   "p90": 9332,
+   "p99": 15314,
+   "max": 16383,
+   "mean": 4956.7,
+   "sum": 283934425
+  },
+  "action_prediction": {
+   "n": 31050,
+   "min": 891,
+   "p25": 2145,
+   "p50": 3056,
+   "p75": 4468,
+   "p90": 6902,
+   "p99": 13229,
+   "max": 16371,
+   "mean": 3713.0,
+   "sum": 115289122
+  },
+  "operation": {
+   "n": 31050,
+   "min": 891,
+   "p25": 2145,
+   "p50": 3056,
+   "p75": 4468,
+   "p90": 6902,
+   "p99": 13229,
+   "max": 16371,
+   "mean": 3713.0,
+   "sum": 115289122
+  },
+  "text": {
+   "n": 6839,
+   "min": 77,
+   "p25": 3887,
+   "p50": 6102,
+   "p75": 8821,
+   "p90": 11587,
+   "p99": 15489,
+   "max": 16376,
+   "mean": 6380.8,
+   "sum": 43638584
+  },
+  "type_target": {
+   "n": 1145,
+   "min": 77,
+   "p25": 128,
+   "p50": 203,
+   "p75": 4354,
+   "p90": 5543,
+   "p99": 10689,
+   "max": 16364,
+   "mean": 2339.1,
+   "sum": 2678228
+  },
+  "select_target": {
+   "n": 265,
+   "min": 83,
+   "p25": 4992,
+   "p50": 5556,
+   "p75": 8341,
+   "p90": 13549,
+   "p99": 16260,
+   "max": 16378,
+   "mean": 6612.7,
+   "sum": 1752365
+  }
+ },
+ "grounding_num_options_bins": {
+  "2-2": 919,
+  "3-5": 946,
+  "6-10": 1090,
+  "11-20": 3239,
+  "21-32": 12060,
+  "33-64": 10501,
+  "65-128": 4307,
+  "129-255": 10
+ },
+ "num_tokens_bins": {
+  "0-1000": 777,
+  "1001-2000": 7397,
+  "2001-4000": 21493,
+  "4001-8000": 23008,
+  "8001-12000": 8342,
+  "12001-16384": 3105
+ },
+ "websites": {
+  "distinct": 1443
+ },
+ "distinct": {
+  "tasks": 3858,
+  "instances": 40230,
+  "tasks_by_source": {
+   "MolmoWeb": 2425,
+   "WebGym": 1433
+  }
+ }
+}