arXiv AI
Aug 6
Item Response Theory for AI Safety
arXiv:2608. 05086v1 Announce Type: new Abstract: Language models differ in how safely they behave and these differences are measured by safety benchmarks.
Efficient Safety Benchmarking via Item Response Theory
The paper demonstrates that Item Response Theory (IRT) can uncover meaningful structure in safety benchmarks for language models, allowing adaptive item selection to approximate full benchmark rankings with Spearman’s ρ > 0.90 while cutting evaluation costs by at least 80% and up to 99.9% on some suites. It also proposes a static method to extract a small, informative subset of items that can be reused across models, achieving 80–99.8% cost savings. These findings show that psychometric techniques can make safety evaluation more efficient without sacrificing ranking accuracy.
Related stories
arXiv AI
Jun 11
Models That Know How Evaluations Are Designed Score Safer
arXiv:2605. 28591v2 Announce Type: replace-cross Abstract: The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings.
arXiv AI
Sep 1
SafeAtlas-VL: Beyond Binary Multimodal Safety with Large-Scale Data and Guard Models
SafeAtlas-VL introduces a large multimodal safety dataset with 1.5 million instances, rating image, request, and response risks on a five‑level ordinal scale across 15 harm categories and 55 subcategories. The accompanying SafeAtlas‑Bench provides 5,000 held‑out cases for evaluating ordinal predictions and continuous risk scores. Models trained on this data, including an 8B Guard model, achieve state‑of‑the‑art performance, outperforming prior benchmarks by about 4% in F1 score.
Hugging Face Trending Papers
Sep 1
Who Judges the Judges? A Chinese Safety QA Benchmark for Evaluating LLM Responses and Safety Judges
Safety benchmarks for large language models often assess the risk of a user query, although the outcome of question answering depends on whether the response violates a policy. This distinction is cri...
arXiv AI
Jul 17
Can We Trust Item Response Theory for AI Evaluation?
arXiv:2607. 15190v1 Announce Type: new Abstract: AI benchmarks increasingly leverage item-level statistical models, particularly item response theory (IRT), to estimate model capabilities, rank systems, select informative examples, and diagnose benchmark quality.
arXiv Machine Learning
Jun 26
DualEval: Joint Model-Item Calibration for Unified LLM Evaluation
arXiv:2606. 26429v1 Announce Type: new Abstract: Current LLM evaluation relies on two complementary but often disconnected signals: static benchmarks with objective correctness labels and arena-style preference data that better reflect open-ended user interactions.
arXiv AI
Aug 19
Benchmarking the Benchmarks: Evaluating Automated Safety Benchmarks for Small Language Models
The paper investigates whether existing AI safety benchmarks, designed for large language models, are suitable for evaluating small language models (SLMs). By testing five benchmark suites on 26 open‑source SLMs with a unified scoring rubric, the authors find that ambiguous judgments dominate, especially for complex prompts and certain architectures. This ambiguity, linked to factors like lexical density and output perplexity, undermines the reliability of aggregate leaderboards and reveals a confound between model capability and perceived safety.
arXiv AI
Jun 8
SafeGene: Reusable Adapters for Transferable Safety Alignment
arXiv:2606. 06519v1 Announce Type: new Abstract: Open-weight LLMs are increasingly fine-tuned into customized assistants, but downstream fine-tuning can weaken safety alignment and make models more vulnerable to malicious prompts, even when the training data is not intentionally harmful.
arXiv AI
Sep 2
Who Judges the Judges? A Chinese Safety QA Benchmark for Evaluating LLM Responses and Safety Judges
arXiv:2609.01210v1 Announce Type: cross Abstract: Safety benchmarks for large language models often assess the risk of a user query, although the outcome of question answering depends on whether the...
arXiv Machine Learning
Sep 17
Safety-Flag: A Unified Benchmark for the Reliability and Calibration of LLM Content Moderators
Safety-Flag is a unified benchmark that consolidates seven popular safety datasets into a single balanced flag/do‑not‑flag protocol, providing item‑level decisions and confidence scores for multiple large language models and dedicated guards. The benchmark evaluates moderator reliability across three dimensions—error direction, probability calibration, and confidence‑based error ranking—revealing that aggregate accuracy masks significant differences, such as one model flagging 85% of benign content while another misses 54% of harmful content. The study shows that general‑purpose models are overconfident, but temperature tuning can substantially improve calibration, and confidence‑based abstention can reduce selective risk, though performance varies with how well confidence ranks errors.
arXiv AI
Aug 12
Why Do Safety Guardrails Degrade Across Languages?
arXiv:2605. 17173v2 Announce Type: replace-cross Abstract: Large language models exhibit safety degradation in non-English languages.
arXiv AI
Jun 16
Adaptive and Explicit safe: Triggering Latent Safety Awareness in Large Reasoning Models
arXiv:2606. 16808v1 Announce Type: new Abstract: While Large Reasoning Models (LRMs) excel at complex tasks, they remain highly vulnerable to sophisticated jailbreaks and direct harmful queries.