Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures
Paper • 2609.29429 • Published • 5
🔄 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
upvoted a paper about 15 hours ago
Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures updated a collection 1 day ago
WTF GENIUS PAPERS upvoted a paper 1 day ago
Six Layers Less: Encoder Pruning for Whisper with Label-Free RecoveryOrganizations
upvoted a paper about 15 hours ago
upvoted 7
papers 1 day ago
Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery
Paper • 2609.27980 • Published • 4
Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models
Paper • 2609.26637 • Published • 15
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
Paper • 2609.29845 • Published • 71
GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression
Paper • 2609.25963 • Published • 13
FLEET: From Logits Entropy to Enhanced Trajectories in Text Generation
Paper • 2609.27657 • Published • 4
Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification Alone
Paper • 2609.23087 • Published • 7
Rufus-Air: An Open LLM Post-Training Recipe
Paper • 2609.29421 • Published • 16
upvoted 4
papers 3 days ago
Hunyuan-A13B Technical Report
Paper • 2609.27284 • Published • 10
MemoryAthena: Adaptive Routing over Latent and Generated Memories
Paper • 2609.25853 • Published • 5
Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms
Paper • 2609.27321 • Published • 12
KITE: KV-Invariant Transformer Expansion for Efficient Agentic LLM Scaling
Paper • 2609.27294 • Published • 1
upvoted 8
papers 4 days ago
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
Paper • 2609.26796 • Published • 32
The Functionalizer: Lossless Functional Decomposition for Subword Tokenization
Paper • 2609.15991 • Published • 17
Towards Full Pipeline FP8 Reinforcement Learning for LLMs
Paper • 2609.22870 • Published • 16
The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks
Paper • 2609.25804 • Published • 152
Complex KDA: Understanding and Enhancing the Expressivity of Kimi Delta Attention
Paper • 2609.24797 • Published • 11
onPanda: Efficient Annotation of On-Policy Alignment Data for LLMs and Agents via Token-Level Correction
Paper • 2609.24983 • Published • 55
Bellman Policy Optimization
Paper • 2609.15987 • Published • 46
LatentPort: Beyond KV Cache - Cross-Model Transfer of Recurrent Memory in Hybrid Language Models: A 4B-to-9B Hybrid-State Handoff Without Target Prefix Replay
Paper • 2609.25053 • Published • 13