# 100 instances is all you need: predicting the success of a new LLM on unseen data by testing on a few instances

arXiv: https://arxiv.org/abs/2409.03563 · HTML: https://mlsanity.com/paper/2409.03563/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 40 references that are in the map.
- First posted on arXiv 5 Sep 2024.

## Builds on (most cited first, 21 of 21)

- [Efficient Estimation of Word Representations in Vector Space](https://mlsanity.com/paper/1301.3781/) (2013)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [Towards Reasoning in Large Language Models: A Survey](https://mlsanity.com/paper/2212.10403/) (2022)
- [Language Models can Solve Computer Tasks](https://mlsanity.com/paper/2303.17491/) (2023)
- [tinyBenchmarks: evaluating LLMs with fewer examples](https://mlsanity.com/paper/2402.14992/) (2024)
- [Abductive Commonsense Reasoning](https://mlsanity.com/paper/1908.05739/) (2019)
- [ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning](https://mlsanity.com/paper/2002.04326/) (2020)
- [Observational Scaling Laws and the Predictability of Language Model Performance](https://mlsanity.com/paper/2405.10938/) (2024)
- [Anchor Points: Benchmarking Models with Much Fewer Examples](https://mlsanity.com/paper/2309.08638/) (2023)
- [Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models -- A Survey](https://mlsanity.com/paper/2404.01869/) (2024)
- [Reasoning Over Paragraph Effects in Situations](https://mlsanity.com/paper/1908.05852/) (2019)
- [MixEval: Deriving Wisdom of the Crowd from LLM Benchmark Mixtures](https://mlsanity.com/paper/2406.06565/) (2024)
- [Revealing the structure of language model capabilities](https://mlsanity.com/paper/2306.10062/) (2023)
- [How predictable is language model benchmark performance?](https://mlsanity.com/paper/2401.04757/) (2024)
- [Towards LogiGLUE: A Brief Survey and A Benchmark for Analyzing Logical Reasoning Capabilities of Language Models](https://mlsanity.com/paper/2310.00836/) (2023)
- [How Predictable Are Large Language Model Capabilities? A Case Study on BIG-bench](https://mlsanity.com/paper/2305.14947/) (2023)
- [Large Language Models for UAVs: Current State and Pathways to the Future](https://mlsanity.com/paper/2405.01745/) (2024)
- [Examining the robustness of LLM evaluation to the distributional assumptions of benchmarks](https://mlsanity.com/paper/2404.16966/) (2024)
- [Predictable Artificial Intelligence](https://mlsanity.com/paper/2310.06167/) (2023)
- [A Challenging Benchmark for Low-Resource Learning](https://mlsanity.com/paper/2303.03840/) (2023)
- [SpaceNLI: Evaluating the Consistency of Predicting Inferences in Space](https://mlsanity.com/paper/2307.02269/) (2023)

## Built on by (most cited first, 24 of 24)

- [Lost in Benchmarks? Rethinking Large Language Model Benchmarking with Item Response Theory](https://mlsanity.com/paper/2505.15055/) (2025)
- [No Answer Needed: Predicting LLM Answer Accuracy from Question-Only Linear Probes](https://mlsanity.com/paper/2509.10625/) (2025)
- [Toward Generalizable Evaluation in the LLM Era: A Survey Beyond Benchmarks](https://mlsanity.com/paper/2504.18838/) (2025)
- [ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities](https://mlsanity.com/paper/2412.06745/) (2024)
- [Beyond One-Size-Fits-All: Tailored Benchmarks for Efficient Evaluation](https://mlsanity.com/paper/2502.13576/) (2025)
- [EffiEval: Efficient and Generalizable Model Evaluation via Capability Coverage Maximization](https://mlsanity.com/paper/2508.09662/) (2025)
- [MiniLongBench: The Low-cost Long Context Understanding Benchmark for Large Language Models](https://mlsanity.com/paper/2505.19959/) (2025)
- [SparseEval: Efficient Evaluation of Large Language Models by Sparse Optimization](https://mlsanity.com/paper/2602.07909/) (2026)
- [PredictaBoard: Benchmarking LLM Score Predictability](https://mlsanity.com/paper/2502.14445/) (2025)
- [PairBench: Are Vision-Language Models Reliable at Comparing What They See?](https://mlsanity.com/paper/2502.15210/) (2025)
- [Can Code-Switched Texts Activate a Knowledge Switch in LLMs? A Case Study on English-Korean Code-Switching](https://mlsanity.com/paper/2410.18436/) (2024)
- [Model Performance-Guided Evaluation Data Selection for Effective Prompt Optimization](https://mlsanity.com/paper/2505.10736/) (2025)
- [Generalized Correctness Models: Learning Calibrated and Model-Agnostic Correctness Predictors from Historical Patterns](https://mlsanity.com/paper/2509.24988/) (2025)
- [On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows](https://mlsanity.com/paper/2605.06110/) (2026)
- [Anticipatory Evaluation of Language Models: Predicting LLM Benchmark Scores from Descriptions](https://mlsanity.com/paper/2509.20645/) (2025)
- [Efficient Benchmarking Is Just Feature Selection and Multiple Regression](https://mlsanity.com/paper/2605.25773/) (2026)
- [Can a Crow Hatch a Falcon? Lineage Matters in Predicting Large Language Model Performance](https://mlsanity.com/paper/2504.19811/) (2025)
- [Capability Salience Vector: Fine-grained Alignment of Loss and Capabilities for Downstream Task Scaling Law](https://mlsanity.com/paper/2506.13216/) (2025)
- [Achieving Trustworthy Real-Time Decision Support Systems with Low-Latency Interpretable AI Models](https://mlsanity.com/paper/2506.20018/) (2025)
- [Fast and Accurate Probing of In-Training LLMs' Downstream Performances](https://mlsanity.com/paper/2604.01025/) (2026)
- [LLM Evaluation on Unseen Questions: Contextual Multidimensional IRT Model](https://mlsanity.com/paper/2608.22295/) (2026)
- [What should an AI assessor optimise for?](https://mlsanity.com/paper/2502.00365/) (2025)
- [Validity Threats for Foundation Model Research](https://mlsanity.com/paper/2606.05029/) (2026)
- [Why Do LLMs Fail at OCL Generation? A Graph Reasoning Perspective](https://mlsanity.com/paper/2609.26122/) (2026)

## Abstract (arXiv)

Predicting the performance of LLMs on individual task instances is essential to ensure their reliability in high-stakes applications. To do so, a possibility is to evaluate the considered LLM on a set of task instances and train an assessor to predict its performance based on features of the instances. However, this approach requires evaluating each new LLM on a sufficiently large set of task instances to train an assessor specific to it. In this work, we leverage the evaluation results of previously tested LLMs to reduce the number of evaluations required to predict the performance of a new LLM. In practice, we propose to test the new LLM on a small set of reference instances and train a generic assessor which predicts the performance of the LLM on an instance based on the performance of the former on the reference set and features of the instance of interest. We conduct empirical studies on HELM-Lite and KindsOfReasoning, a collection of existing reasoning datasets that we introduce, where we evaluate all instruction-fine-tuned OpenAI models until the January 2024 version of GPT4. When predicting performance on instances with the same distribution as those used to train the generic assessor, we find this achieves performance comparable to the LLM-specific assessors trained on the full set of instances. Additionally, we find that randomly selecting the reference instances performs as well as some advanced selection methods we tested. For out of distribution, however, no clear winner emerges and the overall performance is worse, suggesting that the inherent predictability of LLMs is low.