Sharpening Tax in Post-Training
Paper • 2610.01509 • Published • 104
Vivaan Malhotra
vivaan-mal
AI & ML interests
chunking strategies and ColBERT-style retrieval
Recent Activity
liked a dataset about 12 hours ago
yzhuang/Agentic-Long-Context-Understanding-QA liked a model about 12 hours ago
facebook/rag-token-nq upvoted a paper about 12 hours ago
Sharpening Tax in Post-TrainingOrganizations
None yet
upvoted 5
papers about 12 hours ago
World Observer: Joint Actor-Observer Generation for Persistent World Modeling
Paper • 2610.02162 • Published • 85
PhysVista: Benchmarking Physical Intelligence in VLMs via a Perception-Reasoning-Assessment Loop
Paper • 2610.00559 • Published • 30
On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics
Paper • 2609.35259 • Published • 196
OmniSeek: Native Tool Integration for Multi-turn Audio-Visual Reasoning
Paper • 2610.02181 • Published • 21
upvoted 3
papers 1 day ago
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents
Paper • 2609.39102 • Published • 672
Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence
Paper • 2609.34563 • Published • 217
LoopVL: Recurrent Visual Intelligence
Paper • 2609.38426 • Published • 470
upvoted 4
papers 2 days ago
Knowledge Pull Requests for Continual Document Authoring
Paper • 2609.26634 • Published • 12
WorldAttention: An Efficient Attention Architecture for Interactive Video World Models
Paper • 2609.34606 • Published • 49
MemoryAthena: Adaptive Routing over Latent and Generated Memories
Paper • 2609.25853 • Published • 16
SAKI: Maximal-Coupling-Routed Teacher Supervision for On-Policy Distillation
Paper • 2609.36601 • Published • 94
upvoted 4
papers 5 days ago
LongCat-DeepResearch Technical Report
Paper • 2609.36071 • Published • 67
Omni-IO Skills: Harnessing Your Agent Omni-Native
Paper • 2609.31847 • Published • 448
SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Video
Paper • 2609.37969 • Published • 41
EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments?
Paper • 2609.37686 • Published • 70
upvoted 4
papers 7 days ago
D-JEPA: A Decision-Aligned Latent World Model
Paper • 2609.24749 • Published • 17
TimeEvo: Failure-Driven Self-Evolution of a Time Series Agent
Paper • 2609.27277 • Published • 33
Block Sparse Attention with Log-Linear Complexity
Paper • 2609.31093 • Published • 30
MOPD-Router: Rethinking Teacher Routing in Multi-Teacher On-Policy Distillation
Paper • 2609.30837 • Published • 8