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 19

Submitted by

xiaochonglinghu

Video-CoE: Reinforcing Video Event Prediction via Chain of Events

GD-ML

Submitted by

Huaxiu

MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild

UNC-ChapelHill

University of North Carolina at Chapel Hill

Submitted by

ligongh

MosaicMem: Hybrid Spatial Memory for Controllable Video World Models

·
11 authors

3

Submitted by

EilamSha

Alignment Makes Language Models Normative, Not Descriptive

Technion

Technion Israel institute of technology

Submitted by

PumpkinCat

Complementary Reinforcement Learning

alibaba-inc

Submitted by

Jeff-Wang

GigaWorld-Policy: An Efficient Action-Centered World--Action Model

open-gigaai

Submitted by

lyl010221-pku

Look Before Acting: Enhancing Vision Foundation Representations for Vision-Language-Action Models

PekingUniversity

Peking University

2

Submitted by

AIcell

When AI Navigates the Fog of War

umd-zhou-lab

Submitted by

Jungang

Temporal Gains, Spatial Costs: Revisiting Video Fine-Tuning in Multimodal Large Language Models

·
15 authors

Submitted by

niladridutt

LoST: Level of Semantics Tokenization for 3D Shapes

adobe

Submitted by

Sunkyoung

BenchPreS: A Benchmark for Context-Aware Personalized Preference Selectivity of Persistent-Memory LLMs

LGAI-EXAONE

1

Submitted by

Aurumting

ESPIRE: A Diagnostic Benchmark for Embodied Spatial Reasoning of Vision-Language Models

·
5 authors

Submitted by

zawnpn

Conservative Offline Robot Policy Learning via Posterior-Transition Reweighting

BeingBeyond

Submitted by

huanngzh

Stereo World Model: Camera-Guided Stereo Video Generation

·
7 authors

Submitted by

ysner

AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents

·
9 authors

Submitted by

taesiri

Unified Spatio-Temporal Token Scoring for Efficient Video VLMs

allenai

Submitted by

taesiri

LaDe: Unified Multi-Layered Graphic Media Generation and Decomposition

·
3 authors

Submitted by

taesiri

Efficient Exploration at Scale

google

Submitted by

mohammad2012191

VideoAtlas: Navigating Long-Form Video in Logarithmic Compute

KAUST

King Abdullah University of Science and Technology

Submitted by

RaghavvGoel

Efficient Training-Free Multi-Token Prediction via Embedding-Space Probing

qualcomm

Submitted by

ArpitSinghGautam

RAMP: Reinforcement Adaptive Mixed Precision Quantization for Efficient On Device LLM Inference

·
2 authors

Submitted by

xiaorui638

FINER: MLLMs Hallucinate under Fine-grained Negative Queries

TUM

Technical University of Munich

Submitted by

aydnarda

ACE-LoRA: Graph-Attentive Context Enhancement for Parameter-Efficient Adaptation of Medical Vision-Language Models

·
4 authors

Submitted by

taesiri

AI Scientist via Synthetic Task Scaling

·
2 authors

Submitted by

jahid021

HeBA: Heterogeneous Bottleneck Adapters for Robust Vision-Language Models

·
1 authors

Submitted by

taesiri

PRISM: Demystifying Retention and Interaction in Mid-Training

·
4 authors

Submitted by

nstar1125

Coherent Human-Scene Reconstruction from Multi-Person Multi-View Video in a Single Pass

SeoulNatlUniv

Seoul National University

2

Submitted by

gagan3012

Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA

QCRI

Qatar Computing Research Institute