REVIEW 8 cited by
NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
In this position paper, we argue that the classical evaluation on Natural Language Processing (NLP) tasks using annotated benchmarks is in trouble. The worst kind of data contamination happens when a Large Language Model (LLM) is trained on the test split of a benchmark, and then evaluated in the same benchmark. The extent of the problem is unknown, as it is not straightforward to measure. Contamination causes an overestimation of the performance of a contaminated model in a target benchmark and associated task with respect to their non-contaminated counterparts. The consequences can be very harmful, with wrong scientific conclusions being published while other correct ones are discarded. This position paper defines different levels of data contamination and argues for a community effort, including the development of automatic and semi-automatic measures to detect when data from a benchmark was exposed to a model, and suggestions for flagging papers with conclusions that are compromised by data contamination.
Forward citations
Cited by 8 Pith papers
-
Information Discernment in Large Language Models
LLMs update their stated numeric beliefs almost regardless of source reliability or whether a claim moves them closer to the truth, performing near chance on both dimensions.
-
DecompSR: A dataset for decomposed analyses of compositional multihop spatial reasoning
DecompSR is a large, symbolically verified benchmark dataset and generation framework that independently varies productivity, substitutivity, overgeneralisation, and systematicity to probe compositional multihop spati...
-
On the Fitness Landscape in the $NK$ Model
For the NK fitness landscape with K/N tending to alpha, exact limits for free energy and maximum fitness are identified, together with the geometry of near-fittest peaks.
-
Establishing Trustworthy LLM Evaluation via Shortcut Neuron Analysis
Shortcut neuron patching suppresses benchmark-contamination shortcuts in LLMs and yields evaluation scores that strongly correlate with the external MixEval benchmark.
-
RewardAnything: Generalizable Principle-Following Reward Models
RewardAnything follows natural-language reward principles at inference time and, with the new RABench benchmark, demonstrates that principle-conditioned listwise training beats fixed-preference reward models on held-o...
-
Correcting Prompt Dependence in LLM Benchmarks: A Bayesian Hierarchical Model with Embedding-Space Clustering
A Bayesian model that groups similar LLM test prompts into clusters gives better predictive scores than a no-clustering baseline but does not prove that it truly corrects prompt dependence.
-
Unbiased Evaluation of Large Language Models from a Causal Perspective
The paper argues that perturbing benchmark questions with rule-based interventions gives a less contaminated, more interpretable evaluation of LLMs than static benchmarks or agent-generated questions.
-
When Generic Prompt Improvements Hurt: Evaluation-Driven Iteration for LLM Applications
Adding generic prompt rules to task-specific LLM prompts is not monotonic: in 15-20 case local suites, Llama 3 and Qwen 2.5 sometimes pass fewer extraction and RAG checks, so prompt changes should be tested per task.
Discussion (0). Continue with ORCID to comment.