Pith. sign in

REVIEW 1 cited by

Unveiling Safety Vulnerabilities of Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2311.04124 v1 pith:WGGV3ZW7 submitted 2023-11-07 cs.CL cs.AIcs.LG

Unveiling Safety Vulnerabilities of Large Language Models

classification cs.CL cs.AIcs.LG
keywords semanticharmfulmodelmodelsresponsesdatasetidentifyinginappropriate
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

As large language models become more prevalent, their possible harmful or inappropriate responses are a cause for concern. This paper introduces a unique dataset containing adversarial examples in the form of questions, which we call AttaQ, designed to provoke such harmful or inappropriate responses. We assess the efficacy of our dataset by analyzing the vulnerabilities of various models when subjected to it. Additionally, we introduce a novel automatic approach for identifying and naming vulnerable semantic regions - input semantic areas for which the model is likely to produce harmful outputs. This is achieved through the application of specialized clustering techniques that consider both the semantic similarity of the input attacks and the harmfulness of the model's responses. Automatically identifying vulnerable semantic regions enhances the evaluation of model weaknesses, facilitating targeted improvements to its safety mechanisms and overall reliability.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SATORI: Static Test Oracle Generation for REST APIs

    cs.SE 2025-08 unverdicted novelty 5.0

    SATORI statically infers REST API test oracles from OpenAPI specs via LLMs, reporting F1 74.3%, above AGORA+'s 69.3%, with 18 confirmed bugs; the supplied full text, however, is a different paper.