SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2406.07358 · 2024

AI Sandbagging: Language Models can Strategically Underperform on Evaluations

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2406.07358.json · badge: /api/ran/2406.07358.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
teunvdweij/sandbagging canonical (classifier label)
classifier basis: census
5 ran · 0 no recorded run
FunctionStatusWhere it lives
count_correct_answers Ran
ran, honoured the contract we drafted
teunvdweij/sandbagging/plotting/plot_weaker_model_emulation_results.py
pointer only (licence: NONE) · get_code("4130e9405a2f32c9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, teunvdweij/sandbagging 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.

get_base_mc_prompt Ran
ran; our contract draft was wrong, not the code
teunvdweij/sandbagging/src/wmdp_sandbagging/get_benchmark_results_for_hf_models.py
pointer only (licence: NONE) · get_code("a08bd3d753162520")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, teunvdweij/sandbagging 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.

get_sequence_bias Ran
ran; our contract draft was wrong, not the code
teunvdweij/sandbagging/src/wmdp_sandbagging/get_benchmark_results_for_hf_models.py
pointer only (licence: NONE) · get_code("e140939aab3ae1c9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, teunvdweij/sandbagging 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.

init_tokenizer Ran
ran; our contract draft was wrong, not the code
teunvdweij/sandbagging/old/model.py
pointer only (licence: NONE) · get_code("d24e1905ab5edbc9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, teunvdweij/sandbagging 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.

load_sequence_from_jsonl Ran
ran; our contract draft was wrong, not the code
teunvdweij/sandbagging/src/utils.py
pointer only (licence: NONE) · get_code("d332f3065458bbd9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, teunvdweij/sandbagging 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.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 5 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

Trustworthy capability evaluations are crucial for ensuring the safety of AI systems, and are becoming a key component of AI regulation. However, the developers of an AI system, or the AI system itself, may have incentives for evaluations to understate the AI's actual capability. These conflicting interests lead to the problem of sandbagging, which we define as strategic underperformance on an evaluation. In this paper we assess sandbagging capabilities in contemporary language models (LMs). We prompt frontier LMs, like GPT-4 and Claude 3 Opus, to selectively underperform on dangerous capability evaluations, while maintaining performance on general (harmless) capability evaluations. Moreover, we find that models can be fine-tuned, on a synthetic dataset, to hide specific capabilities unless given a password. This behaviour generalizes to high-quality, held-out benchmarks such as WMDP. In addition, we show that both frontier and smaller models can be prompted or password-locked to target specific scores on a capability evaluation. We have mediocre success in password-locking a model to mimic the answers a weaker model would give. Overall, our results suggest that capability evaluations are vulnerable to sandbagging. This vulnerability decreases the trustworthiness of evaluations, and thereby undermines important safety decisions regarding the development and deployment of advanced AI systems.

For agents

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

get_harvested_code_for_paper("2406.07358")
get_code_for_paper("2406.07358")
have("2406.07358")

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

curl https://syntology.ai/api/ran/2406.07358.json

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

[![Syntology run record](https://syntology.ai/api/ran/2406.07358.svg)](https://syntology.ai/paper/2406.07358)

Connect an agent — have() is free.