Papers › Large language model validity via enhanced conformal prediction methods

Large language model validity via enhanced conformal prediction methods

14 Jun 2024arXiv:2406.09714archive 2025-07-28

John J. Cherian, Isaac Gibbs, Emmanuel J. Candès

We develop new conformal inference methods for obtaining validity guarantees on the output of large language models (LLMs). Prior work in conformal language modeling identifies a subset of the text that satisfies a high-probability guarantee of correctness. These methods work by filtering claims from the LLM's original response if a scoring function evaluated on the claim fails to exceed a threshold calibrated via split conformal prediction. Existing methods in this area suffer from two deficiencies. First, the guarantee stated is not conditionally valid. The trustworthiness of the filtering step may vary based on the topic of the response. Second, because the scoring function is imperfect, the filtering step can remove many valuable and accurate claims. We address both of these challenges via two new conformal methods. First, we generalize the conditional conformal procedure of Gibbs et al. (2023) in order to adaptively issue weaker guarantees when they are required to preserve the utility of the output. Second, we show how to systematically improve the quality of the scoring function via a novel algorithm for differentiating through the conditional conformal procedure. We demonstrate the efficacy of our approach on biography and medical question-answering datasets.

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 32 code samples harvested from 3 repositories linked to this paper: 23 ran (of which 0 constructed an object rather than computing a result; 18 with no instrument failure: 3 honoured, 2 violated, 13 with no contract checked; 5 where Syntology's instrument failed) · 9 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 7 ran (7 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 12 samples from 1 repository, 10 ran (7 with no instrument failure; 3 where Syntology's instrument failed); found in paper text by Syntology: 9 samples from 1 repository, 6 ran (4 with no instrument failure; 2 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 between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

jjcherian/conformal-safety officialmentioned in paperMITSyntology: 11 samples, 7 ran (0 instrument-failed) · 4 unverified report
dmis-lab/olaph mentioned on GitHubpytorchSyntology: 12 samples, 10 ran (3 instrument-failed) · 2 unverified report
n4feng/ResponseQualityAssessment mentioned on GitHubSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Response Quality Assessment for Retrieval-Augmented Generation via Conditional... (arXiv:2506.20978; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 1 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

32 samples harvested: 23 ran (of which 0 constructed an object rather than computing a result; 18 with no instrument failure: 3 honoured, 2 violated, 13 with no contract checked; 5 where Syntology's instrument failed) · 9 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.

3ran · honoured contract
2ran · violated contract
3ran · our draft was wrong
2ran; the contract check could not run
13ran
9unverified

Licence: 12 of the 32 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 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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.

jjcherian/conformal-safety official repository 11 samples, 7 ran (0 instrument-failed) · 4 unverified
best_demos in 6 papers src/atomizer.py ran MIT (permissive) · 50ffc3534adfd228 · report fixture: signature query:str/0/str, bm25:BM25Okapi/0/?, demos_sents:list/1/str, k:int/0/int · returned list · level recorded 2026-09-24
compute_conformity_scores src/conformal.py ran · honoured contract MIT (permissive) · c5b53b2caa9e7c67 · report fixture: signature dataset:list/1/dict, scores_list:list/1/ndarray · returned list · level recorded 2026-09-24
conformal_filter src/conformal.py ran MIT (permissive) · 08b581e4f2313f0c · report fixture: signature dataset:list/1/dict, scores_list:list/1/list, thresholds:list/1/float · returned list · level recorded 2026-09-24
find_unique_element src/dataset.py ran MIT (permissive) · f2b316a3259d0a06 · report fixture: signature lst:list/1/int, condition:callable/0/?, approx_index:int/0/int · returned int · level recorded 2026-09-24
generate_annotation_prompt src/query.py ran MIT (permissive) · e970ec6b296d3b38 · report fixture: signature prompt:str/0/str, subclaims:list/1/dict · returned str · level recorded 2026-09-24
generate_merge_prompt src/query.py ran MIT (permissive) · 153977b5ca408d0b · report fixture: signature prompt:str/0/str, subclaims:list/1/dict · returned str · level recorded 2026-09-24
generate_subclaim_prompt src/query.py ran fingerprinted MIT (permissive) · a305f477743aeac3 · report fixture: signature prompt:str/0/str, response:str/0/str · returned str · level recorded 2026-09-24
calibrate_thresholds src/conformal.py unverified MIT (permissive) · 17beb7b9e7b9c708 · report
get_conformal_loss src/prob_model.py unverified MIT (permissive) · 84209d574223e28e · report
get_prompts in 21 papers src/dataset.py unverified MIT (permissive) · 94a610d9319c1840 · report
s3_open src/aws_utils.py unverified MIT (permissive) · 7ab268a78b3e6254 · report
dmis-lab/olaph community (archive-listed) 12 samples, 10 ran (3 instrument-failed) · 2 unverified
ROUGESCORE pdata_collection.py ran · honoured contract no licence file found · pointer only · 0bf59e178503b9e6 · report fixture: signature pred:str/0/str, answer:str/0/str · returned tuple · level recorded 2026-09-24
auto_eval pred_to_sft.py ran no licence file found · pointer only · cd195a4090c8d1a8 · report fixture: signature dictionary:dict/0/?, alpha:float/0/float, beta:float/0/float, gamma:float/0/float · returned str · level recorded 2026-09-24
best_demos in 6 papers Factscore/factscore/atomic_facts.py ran no licence file found · pointer only · 510528f4db1006bc · report fixture: signature query:str/0/str, bm25:BM25Okapi/0/?, demons_sents:list/1/str, k:int/0/int · returned list · level recorded 2026-09-24
eval_score pred_to_preference.py ran no licence file found · pointer only · 96273073405b5dcc · report fixture: signature inst:dict/0/?, alpha:int/0/int, beta:int/0/int, gamma:int/0/int · returned dict · level recorded 2026-09-24
is_invalid_paragraph_ppl in 7 papers Factscore/factscore/abstain_detection.py ran · violated contract no licence file found · pointer only · 8eb846a8e76a79b8 · report fixture: signature text:str/0/str · returned bool · level recorded 2026-09-24
is_invalid_ppl in 7 papers Factscore/factscore/abstain_detection.py ran · violated contract no licence file found · pointer only · 672bd38d830be48d · report fixture: signature text:str/0/str · returned bool · level recorded 2026-09-24
make_inst pred_to_preference.py ran no licence file found · pointer only · a58480b813525951 · report fixture: signature model_name:str/0/str, chosen_idx:int/0/int, reject_idx:int/0/int, inst:dict/0/? · returned dict · level recorded 2026-09-24
normalize_answer in 304 papers Factscore/factscore/atomic_facts.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 6a96435eba311b08 · report fixture: signature s:str/0/str · returned str · level recorded 2026-09-22
remove_citation in 7 papers Factscore/factscore/abstain_detection.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 3635046105ae0fce · report fixture: signature text:str/0/str · returned str · level recorded 2026-09-24
text_to_sentences in 7 papers Factscore/factscore/atomic_facts.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 13867f872a481bd3 · report fixture: signature text:str/0/str · returned list · level recorded 2026-09-24
BERTSCORE pdata_collection.py unverified no licence file found · pointer only · 91e4954df685088b · report
BLEURT pdata_collection.py unverified no licence file found · pointer only · 68d4cb0d0d0e65db · report
jjcherian/conditional-conformal found in paper text by Syntology 9 samples, 6 ran (2 instrument-failed) · 3 unverified
binary_search in 9 papers conditionalconformal/condconf.py ran MIT (permissive) · a9044f6ed0707c0e · report fixture: signature func:callable/0/?, min:int/0/int, max:int/0/int, tol:float/0/float · returned tuple · level recorded 2026-09-24
indicator_matrix conditionalconformal/synthetic_data.py ran; the contract check could not run (RAISES) MIT (permissive) · a43705d36c393243 · report fixture: signature scalar_values:list/1/float, disc:list/1/float · returned ndarray · level recorded 2026-09-24
phi_fn_intercept experiments/synthetic_trial.py ran · honoured contract fingerprinted MIT (permissive) · abe7e77347b17747 · report fixture: signature x:ndarray/2/float · returned ndarray · level recorded 2026-09-24
phi_fn_shifts experiments/synthetic_trial.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) · ca62a7e8fcef8799 · report fixture: signature x:ndarray/1/float · returned ndarray · level recorded 2026-09-24
setup_cvx_problem conditionalconformal/condconf.py ran MIT (permissive) · c738d349dd542153 · report fixture: signature x_calib:ndarray/2/float, scores_calib:ndarray/1/float, phi_calib:ndarray/2/float, infinite_params:dict/0/? · returned Problem · level recorded 2026-09-24
split_dataset in 65 papers conditionalconformal/experiment_utils.py ran MIT (permissive) · 4fe1f68fa8d32424 · report fixture: signature dataset:tuple/1/ndarray, n_test:int/0/int, n_calib:int/0/int, rng:RandomState/0/? · returned tuple · level recorded 2026-09-24
finish_dual_setup conditionalconformal/condconf.py unverified MIT (permissive) · 7040aef1e8bc0075 · report
generate_cqr_data conditionalconformal/synthetic_data.py unverified MIT (permissive) · cb73b72629a8bdb8 · report
torch_ts experiments/temperatureScaling.py unverified MIT (permissive) · 24635b5492ae6ce5 · report

Tasks

Conformal PredictionLanguage ModelingLanguage ModellingLarge Language ModelPredictionQuestion Answering

2 archive task tags without a task page not shown.

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