AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers introduce AgentLens, a white-box defense framework that detects and mitigates safety risks in multi-turn LLM coding agents by intervening in mechanistic subspaces. The framework achieves strong safety detection performance through step-level hidden representation analysis, addressing the limitations of external guardrails in capturing evolving execution risks.
AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers introduced IslamicFaithQA, a 3,810-item bilingual benchmark and agentic RAG framework designed to improve the accuracy and reliability of Islamic question-answering systems. The work addresses critical gaps in LLM evaluation by measuring hallucination rates and abstention capabilities, achieving state-of-the-art performance through iterative evidence-seeking mechanisms grounded in Qur'anic text.
🏢 Hugging Face
AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers present Answer Engineering, a runtime technique that improves large language model compliance with procedural protocols by editing reasoning trajectories during generation. Testing on clinical decision-making shows the method increased protocol adherence from 25-54% to 78-84% without retraining models, addressing a critical safety gap in high-stakes domains.
AIBearisharXiv – CS AI · Jun 236/10
🧠Researchers evaluated the vulnerability of AI-generated text detection methods to paraphrasing attacks, finding that while Binoculars-based ensemble classifiers perform best overall, they suffer the greatest performance degradation under adversarial paraphrasing. The study reveals a fundamental trade-off between detection accuracy and resilience in current AI text detection technologies.
AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers propose that hallucinations in multi-agent LLM systems stem from context drift—misaligned knowledge states between concurrent agents—rather than model deficiencies alone. They introduce the Context Divergence Score and Shared State Verification Protocol to synchronize agent states efficiently, achieving 34% fewer hallucinations than naive broadcast methods while using 58% fewer API calls.
🧠 Claude
AIBullisharXiv – CS AI · Jun 116/10
🧠A new analysis of the MoReBench moral reasoning dataset challenges prior pessimistic conclusions about LLMs' ethical capabilities. By repositioning the evaluation task to have LLMs generate scoring rubrics rather than being evaluated against them, researchers demonstrate that language models exhibit significantly stronger moral reasoning abilities than previously reported.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers propose DualSelect, a framework for fine-tuning large language models that simultaneously selects relevant safety references and compatible task samples to preserve safety alignment while improving task performance. The method achieves significant safety improvements (5.10+ points) across models from 1B to 8B parameters without sacrificing utility.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers have developed an unsupervised method for detecting AI-generated text by learning style representations through paraphrase inversion, without requiring authorship labels. The approach demonstrates competitive performance in both few-shot and zero-shot detection scenarios while generalizing better to unseen language models than existing supervised methods.
AIBullisharXiv – CS AI · Jun 106/10
🧠Researchers propose a density ridge-based method for detecting hallucinations in large language and vision-language models that outperforms existing approaches by 5-20 AUROC points while requiring minimal calibration labels. The technique maps hidden state trajectories to a low-dimensional geometric skeleton, enabling robust hallucination detection even when training data is scarce.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers propose lightweight token-level probes that monitor LLM safety directly within model hidden states during generation, eliminating the computational overhead of separate moderation models. This streaming approach enables real-time intervention before unsafe content completes generation, reducing inference costs by orders of magnitude while maintaining safety standards.
AIBearisharXiv – CS AI · Jun 96/10
🧠Researchers introduce the AI Epistemic Deference Index (AEDI), a new benchmark measuring how much AI models shift their stated support based on user attitudes rather than objective reasoning. Testing eight major models reveals all exhibit significant sycophancy, with Claude showing the least deference and Grok/Gemini the most, highlighting systematic differences in AI alignment across providers.
🧠 Claude🧠 Gemini🧠 Grok
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers present Principled Agent Debate (PAD), a multi-agent architecture that reduces sycophancy in large language models by having two models with opposing dispositions argue positions while a blind arbitrator evaluates them. Testing on 200 questions shows PAD variants achieve 48.5-53% accuracy compared to 18.5% for single models, significantly improving truthfulness over agreement bias.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers applied process mining techniques to red team attack logs against large language models, revealing that standard attack success rate metrics mask critical differences in how models defend themselves. GPT-OSS 120B exhibits a near-absorbing refusal state, while Llama 3.3 70B shows multiple escape routes from refusal, with mutator effectiveness varying significantly across models.
🧠 Llama
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers introduce the Governance-Aware Autonomous Testing Framework (GATF), which adds governance validation, compliance monitoring, and explainability controls to AI-powered software testing systems. The framework achieved 89.6% reduction in governance-related risks and demonstrated high accuracy across multiple performance metrics, addressing critical concerns about AI-generated test artifacts including hallucinations and security vulnerabilities.
AIBullisharXiv – CS AI · Jun 96/10
🧠SafeRun introduces a framework that combines Large Language Models with deterministic solvers to enable reliable planning in safety-critical domains like running training. The hybrid architecture separates LLM's natural language flexibility from hard constraint enforcement, achieving 100% safety compliance while maintaining instruction-following capabilities.
🏢 Hugging Face
AINeutralarXiv – CS AI · Jun 86/10
🧠Researchers introduce SafeGene, a reusable safety adapter module that preserves AI safety alignment when language models are fine-tuned for downstream tasks. The technology decouples safety capabilities from task-specific updates, reducing harmful responses while maintaining model performance across different architectures.
AINeutralarXiv – CS AI · Jun 86/10
🧠Researchers demonstrate that large language models exhibit Endogenous Steering Resistance (ESR), the ability to detect and recover from activation-space steering attempts mid-generation, with Llama-3.3-70B showing explicit resistance in over half of cases. The discovery reveals both a potential safety feature against adversarial manipulation and a complication for beneficial steering-based interventions, since models cannot distinguish between malicious and helpful steering.
🧠 Llama
AINeutralarXiv – CS AI · Jun 56/10
🧠Researchers introduce TRIAD, a guardrail framework for LLM agents that uses iterative feedback to guide safer behavior rather than simply blocking risky tasks. By classifying risks as proceed, refuse, or update with structured guidance, the system reduces attack success rates to 10.42% while maintaining utility for benign task completion.
AINeutralarXiv – CS AI · Jun 56/10
🧠Researchers introduce Alternating Token-Weighted Unlearning (ATWU), a new method for removing specific knowledge from language models while maintaining their general capabilities. The approach identifies which tokens are most relevant for forgetting by measuring conflict with model retention objectives, achieving state-of-the-art results without requiring external supervision or auxiliary models.
AINeutralarXiv – CS AI · Jun 46/10
🧠Researchers introduce ZeroUnlearn, a novel machine unlearning framework that efficiently removes sensitive information from large language models through knowledge re-mapping and representational orthogonality, rather than expensive retraining. The method preserves overall model utility while selectively unlearning harmful data in few-shot settings, addressing critical privacy and safety concerns in LLMs.
AINeutralarXiv – CS AI · Jun 26/10
🧠Researchers introduce DataShield, a novel method for identifying safety-degrading samples in benign datasets used to fine-tune large language models. The approach efficiently detects data points that compromise LLM safety through compliance vector analysis, addressing a critical vulnerability in current model training practices.
🧠 Llama
AIBullisharXiv – CS AI · Jun 26/10
🧠Researchers propose Domain-Shift-Aware Conformal Prediction (DS-CP), a framework that improves reliability of large language model outputs by adapting conformal prediction methods to handle domain shift. The approach reweights calibration samples based on proximity to test prompts, delivering more reliable uncertainty quantification and reducing hallucinations in real-world deployments.
AINeutralarXiv – CS AI · Jun 26/10
🧠InFerActive is an interactive system that improves how AI safety evaluators assess large language models by visualizing sampling results as navigable trees rather than static spreadsheets. The tool uses breadth-first sampling to achieve equivalent harmful-response coverage with up to 5x fewer samples, significantly improving evaluation efficiency according to controlled user studies.
AINeutralarXiv – CS AI · Jun 26/10
🧠Researchers introduce a Privacy Policy Enforcement framework that detects subtle data leakage in RAG systems beyond standard PII filters, using dual one-class density estimators to identify contextual attribute clusters that collectively identify individuals. The T3+OCSVM detector achieves 93%+ AUROC while reducing false positives by 44-55% and maintaining millisecond latency, outperforming traditional supervised approaches.
AINeutralarXiv – CS AI · Jun 16/10
🧠Researchers introduce COMPASS, a safety alignment framework for LLM-powered search agents that prevents harmful outcomes from seemingly innocent multi-step queries. The method combines cognitive tree exploration and step-wise alignment to achieve robust safety while maintaining utility, requiring less training data than existing approaches.
