AI研究论文每日精选

START: Self-taught Reasoner with Tools

Token-Efficient Long Video Understanding for Multimodal LLMs

LLMVoX: Autoregressive Streaming Text-to-Speech Model for Any LLM

EgoLife: Towards Egocentric Life Assistant

LINGOLY-TOO: Disentangling Memorisation from Reasoning with Linguistic Templatisation and Orthographic Obfuscation

Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities

LLM as a Broken Telephone: Iterative Generation Distorts Information

IFIR: A Comprehensive Benchmark for Evaluating Instruction-Following in Expert-Domain Information Retrieval

L^2M: Mutual Information Scaling Law for Long-Context Language Modeling

HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization

FuseChat-3.0: Preference Optimization Meets Heterogeneous Model Fusion

How to Steer LLM Latents for Hallucination Detection?

PokéChamp: an Expert-level Minimax Language Agent

Union of Experts: Adapting Hierarchical Routing to Equivalently Decomposed Transformer

The Best of Both Worlds: Integrating Language Models and Diffusion Models for Video Generation

Identifying Sensitive Weights via Post-quantization Integral

Dedicated Feedback and Edit Models Empower Inference-Time Scaling for Open-Ended General-Domain Tasks

Combining Flow Matching and Transformers for Efficient Solution of Bayesian Inverse Problems

Lost in Literalism: How Supervised Training Shapes Translationese in LLMs

Understanding and Predicting Derailment in Toxic Conversations on GitHub

On the Acquisition of Shared Grammatical Representations in Bilingual Language Models