Emergent Unfaithfulness: How Alignment Training Causes Language Models to Silently Override Task Faithfulness
Research
Research News
New papers, breakthroughs and academic AI research
How Divergence Becomes Decision Flips in Compressed Language Models
Research
Certainty Is Not Just Correctness: Rethinking Token-Level Certainty in LLM Reasoning
Research
DuplexSpeechBench-Document Grounding: Benchmarking Document Grounding and Hallucinations in Voice Agents
Research
CAST: Cost-Aware Speculative Trees from One-Pass Block Drafters
Research
Dissonant ballerinas and crafty carrots: a comparative multi-modal analysis of Italian brain rot
Research
UniBuc at SemEval-2024 Task 2: Tailored Prompting with Solar for Clinical NLI
Research
Refusal Localizes, the Damage Relocates: Safety Layers Under Few-Sample Fine-Tuning
Research
LLM-as-a-Judge for Low-Resource Languages: Adapting Ragas and Comparative Ranking for Romanian
Research
DramaAgent: Agentic Storytelling Video Generation
Research
GPEC: Efficient Pre-LLM Gaussian Process Embedding Correction for Cardiac Video Caption Generation
Research
LENS-GRF: Permutation-Invariant Lesion Evidence Network with Gated Residual Fusion for Acne Severity Grading and Multi-Rater Clinical Oracle Analysis
Research