arXiv AI
Mitigating LLM Over-Refusal via Dynamic Semantic Routing Calibration
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Computation and Language
Sep 23
Mitigating LLM Over-Refusal via Dynamic Semantic Routing Calibratione
The paper addresses the problem of over‑refusal in safety‑aligned large language models, where benign instructions are incorrectly rejected. It identifies that a small set of hypersensitive safety heads in transformer attention misfire on hard‑safe prompts, causing abnormal attention entanglement and high‑entropy routing conflicts that block necessary attention to target entities. To mitigate this, the authors propose Semantic Routing Calibration (SRC), a lightweight, training‑free inference framework that dynamically suppresses these hypersensitive heads and fuses logits from dual branches to restore trustworthy reasoning while preserving intrinsic safety performance.
arXiv Machine Learning
Jun 2
Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
arXiv:2606. 00686v1 Announce Type: new Abstract: The prevailing paradigm in large language model (LLM) alignment operates via erasure, filtering unsafe data or training models to strictly refuse harmful prompts.
arXiv Machine Learning
Aug 14
HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models
arXiv:2608. 12821v1 Announce Type: new Abstract: Large language models (LLMs) remain vulnerable to harmful requests and jailbreak attacks.
arXiv Computation and Language
Aug 25
RASET: Router-Agnostic Safety-Critical Expert Tuning Exposes Localized Safety Enforcement Failures in Mixture-of-Experts LLMs
The paper introduces RASET, a router‑agnostic safety‑critical expert tuning framework for Mixture‑of‑Experts (MoE) large language models. RASET identifies a small subset of experts that are responsible for safety enforcement and applies parameter‑efficient tuning only to those experts, preserving the model’s intrinsic routing behavior. Experiments on five open‑weight MoE backbones show that RASET achieves a high safety‑bypass yield, outperforming existing baselines by a significant margin.
Hugging Face Trending Papers
Jul 2
Safety Targeted Embedding Exploit via Refinement
Safety training for large language models (LLMs) is conducted predominantly in English, leaving uncertain how well safety mechanisms generalize to low-resource languages and mixed-language code-switching. We show that this creates an epistemic gap in which models confidently generate harmful responses for inputs that fall outside the distribution of their safety training.
arXiv AI
Sep 2
From Detection to Refusal: Safer LLMs via Circuit-Guided Weight Scaling
arXiv:2609.00051v1 Announce Type: cross Abstract: Despite extensive alignment efforts, Large Language Models (LLMs) remain vulnerable to generating unsafe content under adversarial prompting, yet the...