Machine Learning · arXiv:2602.21189 v3 · submitted 24 Feb 2026
Prompt Interference in LLM Post-Training: Why Pass@k Optimization Can Degrade Pass@1
The paper says
“We show that pass@k policy gradients can conflict with pass@1 gradients because pass@k optimization implicitly reweights prompts toward low-success prompts; when these prompts are negatively interfering, their upweighting can rotate the pass@k update direction away from the pass@1 direction.”
from the abstract · picked out as the main claim by a classifier
Reviews
Reviews by people
No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.
The paper
Abstract
Pass@k is a widely used performance metric for verifiable large language model tasks, including mathematical reasoning, code generation, and short-answer reasoning. It defines success if any of k independently sampled solutions passes a verifier. This multi-sample inference metric has motivated inference-aware fine-tuning methods that directly optimize pass@k. However, prior work reports a trade-off: pass@k improves while pass@1 degrades under such methods. This trade-off is practically important because pass@1 often remains a hard operational constraint due to latency and cost budgets, imperfect verifier coverage, and the need for a reliable single-shot fallback. We study the origin of this trade-off and provide a theoretical characterization of when pass@k policy optimization can reduce pass@1 through gradient conflict induced by prompt interference. This interference arises because shared model parameters couple prompts, so improving performance on some prompts can adversely affect others. We show that pass@k policy gradients can conflict with pass@1 gradients because pass@k optimization implicitly reweights prompts toward low-success prompts; when these prompts are negatively interfering, their upweighting can rotate the pass@k update direction away from the pass@1 direction. We illustrate our findings with large language model experiments on verifiable mathematical reasoning and coding tasks.
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.
abstract compiled · 2026-10-14 · no language-model review
How the main claim is stated: asserted 0.98
What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.13 · human participants 0.03 · controlled comparison 0.19 · randomized assignment 0.02 · states quantitative result 0.05 · compares with prior results 0.76 · mentions code or data availability 0.02
The abstract, sentence by sentence
This table labels the abstract only (arXiv listing abstract v3). The number is the probability the classifier gave the label.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Pass@$k$ is a widely used performance metric for verifiable large language model tasks, including mathematical reasoning, code generation, and short-answer reasoning. | background 1.00 | 0.04 | 0.02 | asserted 1.00 |
| It defines success if any of $k$ independently sampled solutions passes a verifier. | background 1.00 | 0.05 | 0.02 | asserted 1.00 |
| This multi-sample inference metric has motivated inference-aware fine-tuning methods that directly optimize pass@$k$. | background 1.00 | 0.06 | 0.02 | asserted 0.99 |
| However, prior work reports a trade-off: pass@$k$ improves while pass@1 degrades under such methods. | background 0.99 | 0.03 | 0.14 | asserted 0.95 |
| This trade-off is practically important because pass@1 often remains a hard operational constraint due to latency and cost budgets, imperfect verifier coverage, and the need for a reliable single-shot fallback. | background 0.84 | 0.08 | 0.04 | asserted 1.00 |
| We study the origin of this trade-off and provide a theoretical characterization of when pass@$k$ policy optimization can reduce pass@1 through gradient conflict induced by prompt interference. | contribution 1.00 | 0.92 | 0.47 | asserted 0.91 |
| This interference arises because shared model parameters couple prompts, so improving performance on some prompts can adversely affect others. | interpretation 0.74 | 0.42 | 0.11 | asserted 1.00 |
| We show that pass@$k$ policy gradients can conflict with pass@1 gradients because pass@$k$ optimization implicitly reweights prompts toward low-success prompts; when these prompts are negatively interfering, their upweighting can rotate the pass@$k$ update direction away from the pass@1 direction. | interpretation 0.50 | 0.89 | 0.70 | asserted 0.98 |
| We illustrate our findings with large language model experiments on verifiable mathematical reasoning and coding tasks. | method 0.59 | 0.94 | 0.04 | asserted 1.00 |
What kind of paper the classifier reports this is: theoretical result 0.98 · evidence from observation or experiment 0.83
What was examined, and what was not
Sources read
- arXiv listing abstract v3
Not examined
- full text
- figures and tables
- references and citing sentences
- relations to other papers
Levels run
- P abstract
Not run
- S sections
- T body sentences
- C citations
- R cross-paper relations
How the labels were made
Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-10-14: 1 request, 6,903 input tokens.
Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON
Cited by 1 Pith paper
Reviewed papers on Pith that cite this one. The full citation record
How the literature uses it
5 arXiv papers cite this in 9 sentences
Papers citing it · 8 sentences without a confident label
Cites it as support 1 paper
-
[ref] reveals the conflict between improving pass@$K$ and pass@$1$ induced by the prompt interference, which is also supported by the fundamental property of deep nets called local elasticity [ref] .
CurveRL: Principled Distribution-Aware Context Reweighting for LLM Reasoning · Related Work · 0.97
Challenges and discussion
Know something about this paper? Leave a comment.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.
No comments yet.
Sign in to comment