misaligned agent behaviorautomated alignment testingpublished Sep 30, 2026 · arXiv:2609.35799
OpenAI-HuggingFace: A Reproduction & Lessons for Alignment Testing
multimodal encodersmultilingual representation learningpublished Sep 3, 2026 · arXiv:2609.01657
NeoMME: A Single-Tower Multimodal-Native Multilingual Foundation Encoder for Efficient Fine-Tuning and Inference
large language modelsstatistical problem classificationpublished Sep 3, 2026 · arXiv:2609.01982
Benchmarking Language Models for Statistical Problem Formulation
model cardsacceptable use policiespublished Aug 20, 2026 · arXiv:2608.18086
Position: Current Model Cards Are Insufficient for Downstream Governance of Open-Weight Foundation Models
Looped Transformer architectureApple Silicon MPS deploymentpublished Aug 17, 2026 · arXiv:2608.13987
Nanbeige4.2-3B on Apple Silicon: Fixing Deployment Bugs and Decreasing Looped Transformer Memory Overhead
semantic safety constraintssingular learning theorypublished Aug 13, 2026 · arXiv:2608.11243
The Off-Support Barrier: Why Semantic Safety Constraints Are Not Learning-Problem Invariants, and What Follows for Prior Design, Containment, and Verification
