🫨 hey

khtsly

khtsly

·

AI & ML interests

None yet

Recent Activity

updated a dataset 6 minutes ago

khtsly/cache_stage1_medium_merged

published a dataset 6 minutes ago

khtsly/cache_stage1_medium_merged

updated a dataset 13 minutes ago

khtsly/cache_stage1_small_merged

View all activity

Organizations

None yet

upvoted 4 papers about 15 hours ago

RISE: Recursive Improvement via Self-Extrapolating Policy Distillation

Paper • 2609.05295 • Published 4 days ago • 9

Group Adaptive Clipping Policy Optimization

Paper • 2609.00444 • Published 8 days ago • 5

MaxKernel: Agentic Kernel Generation for TPUs

Paper • 2609.04523 • Published 5 days ago • 9

Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference

Paper • 2609.05275 • Published 4 days ago • 13

upvoted a paper 1 day ago

MDN: Parallelizing Stepwise Momentum for Delta Linear Attention

Paper • 2605.05838 • Published May 7 • 6

upvoted a paper 3 days ago

Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning

Paper • 2609.03430 • Published 5 days ago • 166

upvoted 3 papers 4 days ago

Why Gated DeltaNet Survives 4-Bit Quantization: NVFP4 W4A4 for the Recurrent Half of a Hybrid 27B LLM

Paper • 2609.04098 • Published 5 days ago • 74

Rethinking On-Policy Distillation of Large Language Models II: One Training Example

Paper • 2609.04172 • Published 5 days ago • 81

Post-Training Language Models for Gold-Medal Performance in Coding Competitions

Paper • 2609.02849 • Published 6 days ago • 9

upvoted 3 papers 5 days ago

Language Models Can Control Their Own Attention

Paper • 2609.02737 • Published 6 days ago • 64

Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Model

Paper • 2607.22083 • Published Jul 27 • 10

StudentSim: Training LLM-based Student Simulators

Paper • 2609.01591 • Published 7 days ago • 483

upvoted 3 papers 6 days ago

On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability

Paper • 2608.30320 • Published 8 days ago • 53

Normalized Low-Rank Adaptation

Paper • 2608.31036 • Published 8 days ago • 50

Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement

Paper • 2608.31046 • Published 8 days ago • 143

upvoted 2 papers 8 days ago

Self-Improving Pretraining: using post-trained models to pretrain better models

Paper • 2601.21343 • Published Jan 29 • 21

A Programming Paradigm for Spatiotemporal Composability

Paper • 2608.25512 • Published 13 days ago • 16

upvoted 3 papers 10 days ago

D^3-MOPD: Adaptive Dynamic Domain ScheDuling for Efficient Multi-Teacher Distillation

Paper • 2608.24987 • Published 14 days ago • 27

Agent-G^2: Gaussian Guidance for Agentic Reinforcement Learning

Paper • 2608.23318 • Published 15 days ago • 32

TTPO: Test-Time Policy Optimization

Paper • 2608.27448 • Published 12 days ago • 83