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

Mar 11

Submitted by

exander

Geometry-Guided Reinforcement Learning for Multi-view Consistent 3D Scene Editing

GD-ML

Submitted by

BradyFU

Omni-Diffusion: Unified Multimodal Understanding and Generation with Masked Discrete Diffusion

nanjinguniv

Nanjing University

Submitted by

zorik

Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs

google

1

Submitted by

zli12321

MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data

nvidia

Submitted by

taesiri

InternVL-U: Democratizing Unified Multimodal Models for Understanding, Reasoning, Generation and Editing

·
29 authors

Submitted by

Charlie019

Stepping VLMs onto the Court: Benchmarking Spatial Intelligence in Sports

·
14 authors

Submitted by

KaiserWhoLearns

Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs

JohnsHopkins

Johns Hopkins University

Submitted by

netag

Are Audio-Language Models Listening? Audio-Specialist Heads for Adaptive Audio Steering

barilan

Bar-Ilan University

Submitted by

kimmk135

VLM-SubtleBench: How Far Are VLMs from Human-Level Subtle Comparative Reasoning?

KRAFTON

Submitted by

spapi

Do What I Say: A Spoken Prompt Dataset for Instruction-Following

·
8 authors

Submitted by

taesiri

MiniAppBench: Evaluating the Shift from Text to Interactive HTML Responses in LLM-Powered Assistants

·
6 authors

Submitted by

taesiri

Fish Audio S2 Technical Report

fishaudio

Submitted by

BrandonLiu

Streaming Autoregressive Video Generation via Diagonal Distillation

CUHK-CSE

The Chinese University of Hong Kong

Submitted by

Bowieee

Decoupling Reasoning and Confidence: Resurrecting Calibration in Reinforcement Learning from Verifiable Rewards

ICIP

Submitted by

amanchadha

The Reasoning Trap -- Logical Reasoning as a Mechanistic Pathway to Situational Awareness

·
4 authors

1

Submitted by

CosmoSantoni

Compiler-First State Space Duality and Portable O(1) Autoregressive Caching for Inference

ImperialCollegeLondon

Imperial College London

Submitted by

RaghavvGoel

ConFu: Contemplate the Future for Better Speculative Sampling

qualcomm

Submitted by

amanchadha

SAHOO: Safeguarded Alignment for High-Order Optimization Objectives in Recursive Self-Improvement

·
4 authors

1

Submitted by

juyongjiang

ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement Learning

·
6 authors

Submitted by

ZihaoZhu

BrandFusion: A Multi-Agent Framework for Seamless Brand Integration in Text-to-Video Generation

CUHKSZ

Chinese University of Hong Kong, Shenzhen

Submitted by

taesiri

Towards a Neural Debugger for Python

metaresearch

Submitted by

chendelong

Reward Prediction with Factorized World States

·
7 authors

Submitted by

taesiri

A Text-Native Interface for Generative Video Authoring

adobe