Training, learning and inference: unified dynamics of neural systems
Paper • 2608.20965 • Published • 2
Dipankar Sarkar PRO
dipankarsarkar
AI & ML interests
Building the AI-native stack. Agents as infrastructure, safety as architecture, performance as plumbing. I publish the receipts: papers, datasets, demos.
Recent Activity
upvoted a paper about 1 hour ago
Training, learning and inference: unified dynamics of neural systems upvoted a paper about 1 hour ago
J-Zero: Unified Challenger--Solver--Judge Co-Evolution from Zero DataOrganizations
upvoted 2
papers about 1 hour ago
upvoted 5
papers about 2 hours ago
Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning
Paper • 2608.27549 • Published • 20
Rubric-to-Code Credit Assignment for Reinforcement Learning
Paper • 2608.27906 • Published • 3
Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledge
Paper • 2608.28478 • Published • 11
LMSM: LLM Security Framework Inspired by Linux Security Modules
Paper • 2608.25697 • Published • 3
ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL
Paper • 2608.28476 • Published • 16
upvoted 2
papers 1 day ago
upvoted 11
papers 2 days ago
Giga-Embeddings: Mixture-of-Experts Encoders for High-Throughput Text Embeddings
Paper • 2608.23806 • Published • 2
Gated Activation Steering for Reducing Sycophancy & Hallucination in Medical Question Answering
Paper • 2608.23666 • Published • 1
Tunable Tool-Call Rates in LLM Agents via Representation Steering
Paper • 2608.25198 • Published • 1
Plans You Can Check: Verifier-Grounded Learning of an Open-Weight Planner for Executable Video-Editing
Paper • 2608.25622 • Published • 1
RTPO: Reverse-Turn Policy Optimization for Stabilizing Agentic RL Training
Paper • 2608.18682 • Published • 2
Semantic Overlays: Mitigating Prompt Injection with Annotations Beyond Tokens and Steering Vectors
Paper • 2608.23873 • Published • 1
RecurSE: Bounded Recursive Self-Evaluation for LLM Rubric Judges
Paper • 2608.24231 • Published • 1
TacForcing: Streaming Action Generation with Execution-Time Tactile Feedback
Paper • 2608.25798 • Published • 5
Zero-WAM: In-Context World-Action Modeling from Human Videos for Open-Ended Task Generalization
Paper • 2608.26103 • Published • 19
From Seeing to Acting: Smart Glasses as First-Person Intelligence Platforms
Paper • 2608.24877 • Published • 10
Super Star: Towards Streaming Real-time Interactive Agents for Digital Humans
Paper • 2608.24909 • Published • 5