new

Get trending papers in your email inbox once a day!

Get trending papers in your email inbox!

Daily Papers

byAK and the research community

Sep 28

Submitted by

Hongyang-Du

FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders

USC-PSI-Lab

USC Physical Superintelligence (PSI) Lab

Submitted by

BlackSamorez

Disaggregated Quantization: Specializing LLM Prefill and Decode

ISTA-DASLab

IST Austria Distributed Algorithms and Systems Lab

Submitted by

Sunnyhaze

RayOrch: Programming and Executing Lineage-Controlled Multi-Grain Dataflows for Foundation-Model Data Preparation

PekingUniversity

Peking University

Submitted by

Aphelios-Tang

Block Sparse Attention with Log-Linear Complexity

ByteDance-Seed

2

Submitted by

taesiri

InternW0-Δ: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data

·
48 authors

Submitted by

Muyiiiii

TimeEvo: Failure-Driven Self-Evolution of a Time Series Agent

visainc

Submitted by

Muyiiiii

LastOPD: Taming Collapse in Latent On-Policy Distillation

visainc

Submitted by

LizaKovtun

Tactile-JEPA: Topology-Aware Self-Supervised Representation Learning for Distributed Tactile Sensors

·
4 authors

Submitted by

JHLew

FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance

SeoulNatlUniv

Seoul National University

Submitted by

aLocks

Enhancing Photogrammetric Digital Surface Models with Pretrained Diffusion Models and Multimodal Conditioning

thalesgroup

2

Submitted by

linggm

Jev in the Wild: A Data-Driven Analysis of the Jev Model's Functionality, Applications and Ecosystem

CUHK-CSE

The Chinese University of Hong Kong

Submitted by

ayushjain1144

TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene Representations

CarnegieMellonU

Carnegie Mellon University

Submitted by

taesiri

AgentWorld: Benchmarking Long-Horizon Collaboration of Multi-agent LLMs

·
10 authors

Submitted by

YanZhanPKU

SLCA-GRPO: Resolving Cross-Segment Credit Misattribution in Tool-Calling RL

tencent

Submitted by

taesiri

Game Arena: Strategic LLM Evaluation in Competitive Environments

kaggle

Submitted by

xyzeng2000

SAGE: Mitigating Long-Horizon Reasoning Biases via Topological Guidance

·
4 authors

Submitted by

suvradip11

IndicBankBench: Evaluating Safety and Reliability of Language Model Assistants in Indian Retail Banking

NPCI

National Payments Corporation of India

Submitted by

AndreiArhire

LightMIS: Ultra-Lightweight Medical Image Segmentation Without a Stage-Wise Decoder

·
3 authors

Submitted by

hulehule

CARD: Cluster-level Adaptation with Reward-guided Decoding for Personalized Text Generation

ucirvine

2

Submitted by

hulehule

Do Implicit Personalization and Explicit Styles Conflict? PsPLUG: A Lightweight Plug-in for Balancing Personalization and Style in Customized LLMs

ucirvine

1

Submitted by

tsadja

Morphometric Imitation: From Morphology and Contact Aware Hand Retargeting to Sim-to-Real Visuomotor Policy

Berkeley

Submitted by

Shuaijun

D-JEPA: A Decision-Aligned Latent World Model

NEBULIS-Lab

Submitted by

asimkadav

Softmax Reparameterization for Output-Head Quantization

adobe

2

Submitted by

TianzeTurle

MOPD-Router: Rethinking Teacher Routing in Multi-Teacher On-Policy Distillation

GAIR

Submitted by

scylearning

VLA-Precision: Asymmetric Co-Bootstrapping for Efficient Real-World Online RL of Vision-Language-Action Models

ustc

university of science and technology of china

Submitted by

tianchez

TRACE: Temporal Audit and Condition-aware Evaluation of Streaming Video Understanding

omlab

Submitted by

AvinashAmballa

Not All Ranks Are Equal: Budget-Aware LoRA Merging Across Tasks

SamsungResearch

Samsung Research

2

Submitted by

yonghongzhang

Evidence-Grounded Auditing of Identification Assumptions in Climate-Policy Causal Evaluations

·
4 authors

Submitted by

KristianS7

Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching

TUM

Technical University of Munich

Submitted by

Geralt-Targaryen

ZooWork-ShopRanker: An Open, Preference-Aligned E-Commerce Reranker

·
3 authors

Submitted by

andreagurioli1995

CodeGraph: Open-Taxonomy Knowledge Graph for Source Code with Wikidata Grounding

·
5 authors

Submitted by

junshim

BoundInk: Boundary-Aware Online Handwriting Generation

SUNGKYUNARCH

Sungkyunkwan University

Submitted by

ShuyangXiang

Paragraph Boundaries Are Not White Space:Compression Depth as the Signature of Hierarchical Structure

·
1 authors