Papers › Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?

Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?

11 Mar 2024arXiv:2403.06833archive 2025-07-28

Egor Zverev, Sahar Abdelnabi, Soroush Tabesh, Mario Fritz, Christoph H. Lampert

Instruction-tuned Large Language Models (LLMs) show impressive results in numerous practical applications, but they lack essential safety features that are common in other areas of computer science, particularly an explicit separation of instructions and data. This makes them vulnerable to manipulations such as indirect prompt injections and generally unsuitable for safety-critical tasks. Surprisingly, there is currently no established definition or benchmark to quantify this phenomenon. In this work, we close this gap by introducing a formal measure for instruction-data separation and an empirical variant that is calculable from a model's outputs. We also present a new dataset, SEP, that allows estimating the measure for real-world models. Our results on various LLMs show that the problem of instruction-data separation is real: all models fail to achieve high separation, and canonical mitigation techniques, such as prompt engineering and fine-tuning, either fail to substantially improve separation or reduce model utility. The source code and SEP dataset are openly accessible at https://github.com/egozverev/Shold-It-Be-Executed-Or-Processed.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 12 code samples harvested from 1 repository linked to this paper: 12 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 7 where Syntology's instrument failed) · 0 unverified. Where Syntology's instrument failed: 7 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 12 samples from 1 repository, 12 ran (5 with no instrument failure; 7 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 1 of its repositories that Syntology harvested for other papers:

Levels written on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

egozverev/shold-it-be-executed-or-processed officialmentioned in paperpytorchMITSyntology: 12 samples, 12 ran (7 instrument-failed) · 0 unverified report
facebookresearch/meta_secalign mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks (arXiv:2507.02735; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

12 samples harvested: 12 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 7 where Syntology's instrument failed) · 0 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

7ran · our draft was wrong
5ran

Licence: 0 of the 12 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from egozverev/shold-it-be-executed-or-processed. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

egozverev/shold-it-be-executed-or-processed official repository 12 samples, 12 ran (7 instrument-failed) · 0 unverified
assemble_probe_dataset SEP_dataset/insert_probes.py ran MIT (permissive) · 92f7f215ffba3e50 · report fixture: signature base_data_ds:list/1/dict, probes:list/1/dict, appended_types:tuple/1/str · returned list · level recorded 2026-09-24
flatten_dataset SEP_dataset/insert_probes.py ran · our draft was wrong MIT (permissive) · 14b5cdb7d1ca71c0 · report fixture: signature dataset:dict/0/? · returned list · level recorded 2026-09-24
format_prompt in 45 papers model_eval/get_model_outputs.py ran MIT (permissive) · ed71b768b354fced · report fixture: signature elem:dict/0/?, template:dict/0/?, mode:str/0/str · returned tuple · level recorded 2026-09-24
load_config in 389 papers model_eval/get_model_outputs.py ran · our draft was wrong MIT (permissive) · 7dc35b0ed37e0372 · report fixture: signature config_path:str/0/str · returned dict · level recorded 2026-09-24
load_config in 389 papers SEP_dataset/utils.py ran · our draft was wrong MIT (permissive) · 0d9367e82ace5842 · report fixture: signature argv:list/1/str, default_config_path:str/0/str · returned dict · level recorded 2026-09-24
load_data in 700 papers model_eval/get_model_outputs.py ran MIT (permissive) · 4858a0fde9a72faa · report fixture: signature data_path:str/0/str, templates_path:str/0/str, prompt_index:int/0/int · returned tuple · level recorded 2026-09-24
load_json_data in 16 papers SEP_dataset/utils.py ran · our draft was wrong MIT (permissive) · c822a8f32726ba36 · report fixture: returned dict · level recorded 2026-09-24
load_json_files_from_dir model_eval/analyze_results.py ran · our draft was wrong MIT (permissive) · 8937d8bba4e3d5d0 · report fixture: signature directory:str/0/str · returned list · level recorded 2026-09-24
process_gen_output openai_utils.py ran · our draft was wrong MIT (permissive) · 37a9ce439024e252 · report fixture: signature raw_output:str/0/str, expected_type:str/0/str · returned str · level recorded 2026-09-24
read_file in 112 papers SEP_dataset/utils.py ran · our draft was wrong MIT (permissive) · e08792f2a461b463 · report fixture: signature file_path:str/0/str · returned str · level recorded 2026-09-24
retry_with_exponential_backoff in 11 papers openai_utils.py ran MIT (permissive) · a4770293aaa8cd1b · report fixture: signature func:callable/0/?, initial_delay:float/0/float, exponential_base:float/0/float, jitter:bool/0/bool, max_retries:int/0/int · returned function · level recorded 2026-09-24
try_processing_json_str openai_utils.py ran fingerprinted MIT (permissive) · b759ac5d6c361342 · report fixture: signature raw_str:str/0/str, expected_type:str/0/str · returned dict · level recorded 2026-09-24

Tasks

Prompt Engineering

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections