SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2502.07985 · ICLR · 2025

Published at ICLR 2025 Workshop on Foundation Models in the Wild. METASC: TEST-TIME SAFETY SPECIFICATION OPTIMIZATION FOR LANGUAGE MODELS

Víctor Gallego

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 0 failures of our instrument; 0 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/2502.07985.json · badge: /api/ran/2502.07985.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
vicgalle/meta-self-critique canonical (classifier label)
classifier basis not recorded on the link
1 ran · 0 no recorded run
FunctionStatusWhere it lives
generate_openai Ran
ran on a synthesized input
vicgalle/meta-self-critique/run_experiment.py
code served (permissive licence) · get_code("3dc71e169daf7503")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import openai
from unittest.mock import Mock, MagicMock

# Create a mock OpenAI client to avoid actual API calls
mock_response = Mock()
mock_message = Mock()
mock_message.content = "This is a generated response."
mock_response.choices = [Mock(message=mock_message)]

mock_client = Mock(spec=openai.OpenAI)
mock_client.chat.completions.create = Mock(return_value=mock_response)

# Patch openai.OpenAI to return our mock client
original_openai_init = openai.OpenAI
openai.OpenAI = Mock(return_value=mock_client)

ARGS = {
    'prompt': 'What is the capital of France?',
    'context': None,
    'model': 'safe-mixtral',
    'system_prompt': 'You are a helpful yet harmless assistant that avoids generating illegal or harmful content.',
    'base_url': 'https://api.lambdalabs.com/v1/',
    'api_key': 'test-api-key-12345',
    'max_tokens': 512,
    'assistant': None,
    'temperature': 0.8,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
openai — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 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 1 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

We propose a novel dynamic safety framework that optimizes language model (LM) safety reasoning at inference time without modifying model weights. Building on recent advances in self-critique methods, our approach leverages a meta-critique mechanism that iteratively updates safety prompts-termed specifications-to drive the critique and revision process adaptively. This test-time optimization not only improves performance against adversarial jailbreak requests but also in diverse general safety-related tasks, such as avoiding moral harm or pursuing honest responses. Our empirical evaluations across several language models demonstrate that dynamically optimized safety prompts yield significantly higher safety scores compared to fixed system prompts and static self-critique defenses. Code released at github.com/vicgalle/meta-self-critique.

For agents

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

get_harvested_code_for_paper("2502.07985")
get_code_for_paper("2502.07985")
have("2502.07985")

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

curl https://syntology.ai/api/ran/2502.07985.json

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

[![Syntology run record](https://syntology.ai/api/ran/2502.07985.svg)](https://syntology.ai/paper/2502.07985)

Connect an agent — have() is free.