SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness
Paper • 2609.20519 • Published • 45
🔄 In a Training Loop
Urro
urroxyz
AI & ML interests
computational linguistics major 🤖🔎ðŸ”
i am autistic. if i come off rude, i probably didn't mean to. please feel free to ask me for clarification.
Recent Activity
updated a collection 25 minutes ago
WTF GENIUS PAPERS upvoted a paper 25 minutes ago
SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness updated a collection 25 minutes ago
WTF GENIUS PAPERSOrganizations
upvoted 6
papers 25 minutes ago
When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation
Paper • 2609.20511 • Published • 29
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
Paper • 2609.19969 • Published • 56
What Does Privileged Information Add to On-Policy Self-Distillation?
Paper • 2609.20612 • Published • 4
When2Think: Learning Difficulty-Aware Length Control for Efficient Hybrid Reasoning Models
Paper • 2609.19671 • Published • 4
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
Paper • 2609.20784 • Published • 15
upvoted 7
papers 1 day ago
Continual Learning Mechanisms Compose for Long-Horizon Memorization
Paper • 2609.06986 • Published • 318
ImpossibleRubrics: Stress-Testing Generated Rubrics as Reward Signals
Paper • 2609.16816 • Published • 4
Training Specialist Models without Reasoning Trajectories for Domain Expert Distillation
Paper • 2609.13770 • Published • 3
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence
Paper • 2609.17488 • Published • 167
SpectralShift: Effective Context Window Extension of Gated DeltaNet via Spectral Reparameterization
Paper • 2609.14320 • Published • 23
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
Paper • 2609.18708 • Published • 62
Register Tokens for Bounded-State Reasoning in Diffusion Language Models
Paper • 2609.16372 • Published • 5
upvoted 3
papers 2 days ago
How Lossless Is Lossless Speculative Decoding? The Role of Numerical Precision in Orthrus
Paper • 2609.15504 • Published • 31
Learning to Solve Hard Problems in RL for LLMs by Never Giving Up
Paper • 2609.13443 • Published • 10
Lightning Weave: Improving the Accuracy-Efficiency Frontier of Reasoning Models through Capability Composition
Paper • 2609.14708 • Published • 2
upvoted 4
papers 3 days ago
DataFlex-RL: An Evaluation Platform for RLVR Data Policies
Paper • 2609.06107 • Published • 80
Dream-RSI: Recursive Self-Improvement through Evolving Worlds
Paper • 2609.14858 • Published • 226
Grouped Value Attention: Efficient KV Caching via On-Demand Key Reconstruction
Paper • 2609.13285 • Published • 68
Expert-Space Exploration in MoE Reinforcement Learning
Paper • 2609.13058 • Published • 7