SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.02280 · 2026

RACC: Representation-Aware Coverage Criteria for LLM Safety Testing

Zeming Wei, Yihao Zhang, Chengcan Wu, Zhixin Zhang, Xiaokun Luan, Safety Concepts

arXiv · PDF · Open in the Atlas

Code that ran

7 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (6 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2602.02280.json · badge: /api/ran/2602.02280.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
weizeming/RACC community
Papers with Code archive
7 ran · 1 no recorded run
FunctionStatusWhere it lives
CBC Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("d5c706a6ffcb1027")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

FIC Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("3bcf5a510a246720")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

PCAConceptExtractor Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("d054656a92fb0ec9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

PCC Ran
ran on a synthesized input
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("51da3f3901ad5eac")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

SCC Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("c5b46dcef73836a7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

SFC Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("1c6c7f08551a5aa0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

TKFC Ran
ran on a synthesized input · constructed an object; no result was computed
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("a444d0791440c465")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

compute_racc Not yet run
unverified (harvested, no recorded run)
weizeming/RACC/compute_coverage.py
pointer only (licence: NONE) · get_code("cd4a7098b4a02b55")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, weizeming/RACC when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 8 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Large Language Models (LLMs) face severe safety risks from jailbreak attacks, yet current safety testing largely relies on static datasets and lacks systematic criteria to evaluate test suite quality and adequacy. While coverage criteria have proven effective for smaller neural networks, they are impractical for LLMs due to computational overhead and the entanglement of safety-critical signals with irrelevant neuron activations. To address these issues, we propose RACC (Representation-Aware Coverage Criteria), a set of coverage criteria specialized for LLM safety testing. RACC first extracts safety representations from the LLM's hidden states using a small calibration set of harmful prompts, then measures test prompts' concept activations against these directions, and finally computes coverage through six criteria assessing both individual and compositional safety concept coverage. Experiments on multiple LLMs and safety benchmarks show that RACC reliably rewards high-quality jailbreak test suites while remaining insensitive to redundant or invalid inputs, which is a key distinction that neuron-level criteria fail to make. We further demonstrate RACC's practical value in two applications, including test suite prioritization and attack prompt sampling, and validate its generalization across diverse settings and configurations. Overall, RACC provides a scalable and principled foundation for coverage-guided LLM safety testing.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2602.02280")
get_code_for_paper("2602.02280")
have("2602.02280")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2602.02280.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2602.02280.svg)](https://syntology.ai/paper/2602.02280)

Connect an agent — have() is free.