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 24

Submitted by

HaoboZheng

SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue

zju

Zhejiang University

Submitted by

zwq2018

Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World

OmniAI-ZJU

Submitted by

CheeryLJH

HappyWorld-Bench

alibabagroup

Submitted by

Harold328

The Past Frames the Future: Memory for Autoregressive Video Generation

·
25 authors

Submitted by

yli-ml

Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents

Salesforce

Salesforce AI Research

Submitted by

EddieYang428

RewardVerse: Rubric-Guided Policy Optimization for Video Reward Modeling

tencent

Submitted by

Silin-Chen

Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It?

SJTU

Shanghai Jiao Tong University

Submitted by

Eclipse2001

PACT: From Credit Assignment to Critic Alignment

AllSpark-Research

AllSpark Research

3

Submitted by

Xiaoyuluoit97

Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models

SeaFill2025

2

Submitted by

ammarali32

GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression

MTSAIR

2

Submitted by

donghao-zhou

PackLab: A Comprehensive Framework for Developing, Training, and Evaluating MLLMs in Robotic Bin Packing

CUHK-CSE

The Chinese University of Hong Kong

Submitted by

soujanyaporia

MemBodied: Recurrent Associative Memory for Vision-Language-Action Models

·
7 authors

Submitted by

taesiri

Hunyuan-A13B Technical Report

Tencent-Hunyuan

Tencent Hunyuan

Submitted by

ethanning

WhatWorkedBench: Benchmarking Experimental Understanding in AI Agents

CarnegieMellonU

Carnegie Mellon University

Submitted by

taesiri

Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms

Qwen

Submitted by

ohad204

All modalities are equal, but video is more equal: Closing the Cross-Attention Gap in Joint Video Generation

barilan

Bar-Ilan University

Submitted by

taesiri

InternW0: A Foundational Physical World Model for Efficient Real-World Interactions

·
26 authors

Submitted by

apappu97

Self-Organizing Agent Teams Learn to Reason Together

·
8 authors

Submitted by

chfeng

On the Diffusibility of High-Dimensional Latents

·
10 authors

Submitted by

rasgaard

Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery

·
2 authors

Submitted by

leehenry

MemoryAthena: Adaptive Routing over Latent and Generated Memories

OLAResearchX

Omni Language AI Research

Submitted by

cgnorthcutt

StudentBench: AI and human tutoring yield equivalent GRE learning gains

handshake-ai-research

Handshake AI Research

Submitted by

ruyi101

The Linear Representation Hypothesis Needs a Group Action

·
3 authors

2

Submitted by

alexmartin1722

Knowledge Pull Requests for Continual Document Authoring

·
2 authors

Submitted by

mario-sanz

Calibration as a First-Class Criterion in LLM Evaluation

·
2 authors

2

Submitted by

zeyush

EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics

·
19 authors

Submitted by

wisdompan

X-Planner: Event-Structured Task Planning for Embodied Intelligence

·
32 authors

Submitted by

Alexiush

FLEET: From Logits Entropy to Enhanced Trajectories in Text Generation

·
2 authors

Submitted by

1zeryu

Uranus: Building the Next-Generation Simulation Infrastructure for Embodied AI

D-Robotics