Learning to Read Out: Unembedding Dynamics in Language Model Pretraining
Paper • 2610.21025 • Published • 9
xmufqk
LFQ4XMU
AI & ML interests
None yet
Recent Activity
upvoted a paper about 13 hours ago
Learning to Read Out: Unembedding Dynamics in Language Model Pretraining upvoted a paper about 13 hours ago
Beyond Policy Rollouts: Embedding-Space Directional Exploration with Anchored Noise for LLM RLVR upvoted a paper about 13 hours ago
Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language ModelsOrganizations
None yet
upvoted 9
papers about 13 hours ago
Beyond Policy Rollouts: Embedding-Space Directional Exploration with Anchored Noise for LLM RLVR
Paper • 2610.16105 • Published • 15
Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language Models
Paper • 2610.22428 • Published • 13
Factorization-Guided Training and Inference for Diffusion-Drafted Speculative Decoding
Paper • 2610.23935 • Published • 7
Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight
Paper • 2610.08077 • Published • 190
VideoJev: System One Model for Video Understanding and Interaction
Paper • 2610.18615 • Published • 377
Kimi K3: Open Frontier Intelligence
Paper • 2607.24653 • Published • 534
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence
Paper • 2609.17488 • Published • 803
SpecMRL: Redesign Matryoshka Representation Learning within Parametric Spectrum
Paper • 2610.26824 • Published • 81
upvoted 4
papers 20 days ago
Language Models Are "Insecure" Reporters
Paper • 2609.36139 • Published • 17
VideoLoop: Looped Working Memory Against Semantic Thrashing in Long-Form Video Agents
Paper • 2609.38119 • Published • 12
AutoRef: Harness Optimization for Agentic Multi-Reference Image Generation
Paper • 2609.35530 • Published • 23
Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression
Paper • 2609.36322 • Published • 119
upvoted 6
papers 4 months ago
Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models
Paper • 2606.11324 • Published • 166
ICA Lens: Interpreting Language Models Without Training Another Dictionary
Paper • 2606.11722 • Published • 18
Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning
Paper • 2606.11683 • Published • 31
Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
Paper • 2606.09076 • Published • 67
Redesign Mixture-of-Experts Routers with Manifold Power Iteration
Paper • 2606.12397 • Published • 91
Your UnEmbedding Matrix is Secretly a Feature Lens for Text Embeddings
Paper • 2606.07502 • Published • 100
upvoted a paper 9 months ago