Maxim Kurkin

dondosss's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted a paper 17 days ago

I Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse Autoencoders

Paper • 2503.18878 • Published 18 days ago • 113

upvoted a paper 18 days ago

When Less is Enough: Adaptive Token Reduction for Efficient Image Representation

Paper • 2503.16660 • Published 22 days ago • 71

upvoted a paper 21 days ago

One-Step Residual Shifting Diffusion for Image Super-Resolution via Distillation

Paper • 2503.13358 • Published 25 days ago • 93

upvoted a paper 23 days ago

RWKV-7 "Goose" with Expressive Dynamic State Evolution

Paper • 2503.14456 • Published 24 days ago • 137

upvoted 2 papers about 1 month ago

When an LLM is apprehensive about its answers -- and when its uncertainty is justified

Paper • 2503.01688 • Published Mar 3 • 20

GHOST 2.0: generative high-fidelity one shot transfer of heads

Paper • 2502.18417 • Published Feb 25 • 66

upvoted 12 papers about 2 months ago

LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Token

Paper • 2501.03895 • Published Jan 7 • 53

LLM-Microscope: Uncovering the Hidden Role of Punctuation in Context Memory of Transformers

Paper • 2502.15007 • Published Feb 20 • 172

LightThinker: Thinking Step-by-Step Compression

Paper • 2502.15589 • Published Feb 21 • 28

VLM^2-Bench: A Closer Look at How Well VLMs Implicitly Link Explicit Matching Visual Cues

Paper • 2502.12084 • Published Feb 17 • 29

SurveyX: Academic Survey Automation via Large Language Models

Paper • 2502.14776 • Published Feb 20 • 97

SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines

Paper • 2502.14739 • Published Feb 20 • 102

MLGym: A New Framework and Benchmark for Advancing AI Research Agents

Paper • 2502.14499 • Published Feb 20 • 190

SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

Paper • 2502.14786 • Published Feb 20 • 141

How Much Knowledge Can You Pack into a LoRA Adapter without Harming LLM?

Paper • 2502.14502 • Published Feb 20 • 89

Cramming 1568 Tokens into a Single Vector and Back Again: Exploring the Limits of Embedding Space Capacity

Paper • 2502.13063 • Published Feb 18 • 69

CodeI/O: Condensing Reasoning Patterns via Code Input-Output Prediction

Paper • 2502.07316 • Published Feb 11 • 48

SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators

Paper • 2502.06394 • Published Feb 10 • 90

upvoted a paper 2 months ago

The Differences Between Direct Alignment Algorithms are a Blur

Paper • 2502.01237 • Published Feb 3 • 115

upvoted a collection 2 months ago

Qwen2.5-VL

Collection

Vision-language model series based on Qwen2.5 • 11 items • Updated 11 days ago • 439