On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters
Paper • 2606.02437 • Published • 224
jian
lipliu
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 2 days ago
On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters upvoted a paper 5 days ago
Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models upvoted a paper 6 days ago
Code as Agent HarnessOrganizations
None yet
upvoted a paper 2 days ago
upvoted a paper 5 days ago
upvoted 11
papers 6 days ago
Code as Agent Harness
Paper • 2605.18747 • Published • 215
MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation
Paper • 2605.27366 • Published • 27
Rethinking Memory as Continuously Evolving Connectivity
Paper • 2605.28773 • Published • 34
Agent Explorative Policy Optimization for Multimodal Agentic Reasoning
Paper • 2605.28774 • Published • 90
SkillOpt: Executive Strategy for Self-Evolving Agent Skills
Paper • 2605.23904 • Published • 224
ESPO: Early-Stopping Proximal Policy Optimization
Paper • 2605.29860 • Published • 19
Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories
Paper • 2606.03979 • Published • 29
SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories
Paper • 2606.01311 • Published • 35
Trust Region On-Policy Distillation
Paper • 2606.01249 • Published • 42
Trust-Region Behavior Blending for On-Policy Distillation
Paper • 2605.31159 • Published • 65
Self-Distilled Policy Gradient
Paper • 2606.04036 • Published • 24
upvoted a paper 2 months ago
upvoted 4
papers 3 months ago
Reasoning Models Struggle to Control their Chains of Thought
Paper • 2603.05706 • Published • 39
Weak-Driven Learning: How Weak Agents make Strong Agents Stronger
Paper • 2602.08222 • Published • 290
iGRPO: Self-Feedback-Driven LLM Reasoning
Paper • 2602.09000 • Published • 19
Does Your Reasoning Model Implicitly Know When to Stop Thinking?
Paper • 2602.08354 • Published • 266
upvoted a paper 4 months ago
upvoted a paper 5 months ago