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 7

Submitted by

scyyc9

Bilevel Coordinated Reflection: A Game-Theoretic Approach to Multi-Agent LLM Systems

·
6 authors

Submitted by

circleLZY

Iris: Climbing to the Search Frontier

AllSpark-Research

AllSpark Research

Submitted by

itaowe

Motion-Omni: End-to-End Joint Speech and Full-Body Motion for Spoken Dialogue

PKU1898

Peking University

Submitted by

taesiri

WorldSculpt: Generating Compositional Worlds from Grounded Videos

·
12 authors

Submitted by

SagiPolaczek

The Attention Triangle in Audio-Video Models

tau

Tel Aviv University

Submitted by

julia-bel

Enoki: Efficient Multi-Level Hallucination Detection

s-nlp

Submitted by

Chenyu-Zhou

Ask Before You Optimize: Dynamic Pre-Formulation Clarification for Interactive Optimization

AIOR-Research

Submitted by

taesiri

Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference

cerebras

1

Submitted by

taesiri

MaxKernel: Agentic Kernel Generation for TPUs

google

Submitted by

sg-jeong28

Beneath the Surface of Chains-of-Thought: A Mechanistic Interpretation of Reasoning Operations in LLMs

naver-ai

Submitted by

taesiri

RISE: Recursive Improvement via Self-Extrapolating Policy Distillation

Salesforce

Salesforce AI Research

Submitted by

tyzhu

When Models Edit Too Much: On the Fidelity of Minimal Code Edits

NationalUniversityofSingapore

National University of Singapore

Submitted by

shengjia-toronto

Group Adaptive Clipping Policy Optimization

amazon

Submitted by

ankanmbz

Training-Free Speech-Centric Omni Understanding with Frozen VLMs

MBZUAI

Mohamed Bin Zayed University of Artificial Intelligence

Submitted by

benshi34

τ^τ-Bench: An Environment for End-To-End, Realistic Agent Construction

sierra-research

Sierra Technologies, Inc.

Submitted by

shunpeng

AdaptVPR: Route-Aware Hard Positive Generation for Robust Visual Place Recognition

·
9 authors

Submitted by

Ismailerbas

When Quantization Breaks Memory: Recurrent-State Write-Back in Low-Precision Temporal Inference

·
3 authors

Submitted by

JitaiHao

ShallowStream: Index Shallow then Answer Deep for Streaming Video Understanding

·
4 authors

Submitted by

xB1n

To See a World in a Living Context: Unified Indoor-Outdoor Urban World Generation

·
6 authors