AGI Lab

non-profit

AI & ML interests

None defined yet.

Recent Activity

Ningyu submitted a paper 4 days ago

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning

Ningyu submitted a paper 5 days ago

Rethinking Memory as Continuously Evolving Connectivity

Ningyu submitted a paper 5 days ago

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

View all activity

Papers

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning

Rethinking Memory as Continuously Evolving Connectivity

View all Papers

AGI-LAB-HF 's papers 30

Submitted by

Ningyu Zhang

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning

AGI-LAB-HF

Submitted by

Ningyu Zhang

Rethinking Memory as Continuously Evolving Connectivity

AGI-LAB-HF

2

Submitted by

Ningyu Zhang

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

AGI-LAB-HF

Submitted by

bowen xu

Long Live The Balance: Information Bottleneck Driven Tree-based Policy Optimization

AGI-LAB-HF

Submitted by

xiaochonglinghu

Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos

AGI-LAB-HF

Submitted by

shen

FashionChameleon: Towards Real-Time and Interactive Human-Garment Video Customization

AGI-LAB-HF

Submitted by

Jialiang Cheng

BEAM: Binary Expert Activation Masking for Dynamic Routing in MoE

AGI-LAB-HF

2

Submitted by

zhengrong yue

What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusion

AGI-LAB-HF

Submitted by

liutao

Continuous-Time Distribution Matching for Few-Step Diffusion Distillation

AGI-LAB-HF

Submitted by

xuguo

CoInteract: Physically-Consistent Human-Object Interaction Video Synthesis via Spatially-Structured Co-Generation

AGI-LAB-HF

Submitted by

taesiri

Tstars-Tryon 1.0: Robust and Realistic Virtual Try-On for Diverse Fashion Items

AGI-LAB-HF

7

Submitted by

xiaochonglinghu

Elucidating the SNR-t Bias of Diffusion Probabilistic Models

AGI-LAB-HF

Submitted by

xiaochonglinghu

Omni-WorldBench: Towards a Comprehensive Interaction-Centric Evaluation for World Models

AGI-LAB-HF

Submitted by

Dilxat Muhtar

Complementary Reinforcement Learning

AGI-LAB-HF

Submitted by

bowen xu

MASQuant: Modality-Aware Smoothing Quantization for Multimodal Large Language Models

AGI-LAB-HF

Submitted by

Ningyu Zhang

How Controllable Are Large Language Models? A Unified Evaluation across Behavioral Granularities

AGI-LAB-HF

2

Submitted by

xiaochonglinghu

MobilityBench: A Benchmark for Evaluating Route-Planning Agents in Real-World Mobility Scenarios

AGI-LAB-HF

Submitted by

xiaochonglinghu

From Scale to Speed: Adaptive Test-Time Scaling for Image Editing

AGI-LAB-HF

What if Agents Could Imagine? Reinforcing Open-Vocabulary HOI Comprehension through Generation

AGI-LAB-HF

Submitted by

xiaochonglinghu

FASA: Frequency-aware Sparse Attention

AGI-LAB-HF

Submitted by

bowen xu

D-CORE: Incentivizing Task Decomposition in Large Reasoning Models for Complex Tool Use

AGI-LAB-HF

9

Submitted by

xiaochonglinghu

Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models

AGI-LAB-HF

Submitted by

Litu Ou

Uncertainty-Aware Gradient Signal-to-Noise Data Selection for Instruction Tuning

AGI-LAB-HF

Submitted by

xiaochonglinghu

Urban Socio-Semantic Segmentation with Vision-Language Reasoning

AGI-LAB-HF

Submitted by

xiaochonglinghu

Thinking with Map: Reinforced Parallel Map-Augmented Agent for Geolocalization

AGI-LAB-HF

Submitted by

taesiri

Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem

AGI-LAB-HF

Submitted by

Yang Li (SJTU & SII)

Reasoning Palette: Modulating Reasoning via Latent Contextualization for Controllable Exploration for (V)LMs

AGI-LAB-HF

2

Submitted by

zehan wang

DSI-Bench: A Benchmark for Dynamic Spatial Intelligence

AGI-LAB-HF

Submitted by

Yang Li (SJTU & SII)

Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization

AGI-LAB-HF

2

Submitted by

Wang

Socratic-Zero : Bootstrapping Reasoning via Data-Free Agent Co-evolution

AGI-LAB-HF