English
Related papers

Related papers: Revisiting Methods for Finding Influential Example…

200 papers

Evaluations of large language models (LLMs) suffer from instability, where small changes of random factors such as few-shot examples can lead to drastic fluctuations of scores and even model rankings. Moreover, different LLMs can have…

Machine Learning · Computer Science 2025-09-17 Yiyang Li , Yonghuang Wu , Ying Luo , Liangtai Sun , Zishu Qin , Lin Qiu , Xuezhi Cao , Xunliang Cai

Influence functions (IF) have been seen as a technique for explaining model predictions through the lens of the training data. Their utility is assumed to be in identifying training examples "responsible" for a prediction so that, for…

Machine Learning · Computer Science 2023-05-29 Andrea Schioppa , Katja Filippova , Ivan Titov , Polina Zablotskaia

Language models are commonly fine-tuned via reinforcement learning to alter their behavior or elicit new capabilities. Datasets used for these purposes, and particularly human preference datasets, are often noisy. The relatively small size…

Machine Learning · Computer Science 2025-07-22 Daniel Fein , Gabriela Aranguiz-Dias

As machine learning is increasingly deployed in the real world, it is paramount that we develop the tools necessary to analyze the decision-making of the models we train and deploy to end-users. Recently, researchers have shown that…

Machine Learning · Computer Science 2022-05-05 Andrew Silva , Rohit Chopra , Matthew Gombolay

There is an especially strong need in modern large-scale data analysis to prioritize samples for manual inspection. For example, the inspection could target important mislabeled samples or key vulnerabilities exploitable by an adversarial…

Machine Learning · Statistics 2017-05-11 Mike Wojnowicz , Ben Cruz , Xuan Zhao , Brian Wallace , Matt Wolff , Jay Luan , Caleb Crable

Pretrained Large Language Models (LLMs) achieve strong performance across a wide range of tasks, yet exhibit substantial variability in the various layers' training quality with respect to specific downstream applications, limiting their…

Computation and Language · Computer Science 2025-10-27 Hadi Askari , Shivanshu Gupta , Fei Wang , Anshuman Chhabra , Muhao Chen

Influence functions estimate the effect of removing a training point on a model without the need to retrain. They are based on a first-order Taylor approximation that is guaranteed to be accurate for sufficiently small changes to the model,…

Machine Learning · Computer Science 2019-11-22 Pang Wei Koh , Kai-Siang Ang , Hubert H. K. Teo , Percy Liang

In-context learning (ICL) enables large language models to perform new tasks by conditioning on a sequence of examples. Most prior work reasonably and intuitively assumes that which examples are chosen has a far greater effect on…

Computation and Language · Computer Science 2025-11-14 Warren Li , Yiqian Wang , Zihan Wang , Jingbo Shang

Motivated by the goals of dataset pruning and defect identification, a growing body of methods have been developed to score individual examples within a dataset. These methods, which we call "example difficulty scores", are typically used…

Machine Learning · Computer Science 2024-01-04 Devin Kwok , Nikhil Anand , Jonathan Frankle , Gintare Karolina Dziugaite , David Rolnick

The goal of data attribution is to trace the model's predictions through the learning algorithm and back to its training data. thereby identifying the most influential training samples and understanding how the model's behavior leads to…

Machine Learning · Computer Science 2025-08-12 Hongbo Zhu , Angelo Cangelosi

It is not only what we ask large language models (LLMs) to do that matters, but also how we prompt. Phrases like "This is urgent" or "As your supervisor" can shift model behavior without altering task content. We study this effect as…

Computation and Language · Computer Science 2026-02-26 Yilin Geng , Omri Abend , Eduard Hovy , Lea Frermann

Network meta-analysis is an evidence synthesis method for comparing the effectiveness of multiple available treatments. To justify evidence synthesis, consistency is an important assumption; however, existing methods founded on statistical…

Methodology · Statistics 2025-04-29 Kotaro Sasaki , Hisashi Noma

This paper poses two critical issues in evaluating base models (without post-training): (1) Unstable evaluation during training: in the early stages of pre-training, the models lack the capability to answer questions as required, leading to…

Machine Learning · Computer Science 2025-06-02 Hongzhi Luan , Changxin Tian , Zhaoxin Huan , Xiaolu Zhang , Kunlong Chen , Zhiqiang Zhang , Jun Zhou

Large language models (LLMs) have demonstrated remarkable reasoning capabilities in math and coding, often bolstered by post-training on the chain-of-thoughts (CoTs) generated by stronger models. However, existing strategies for curating…

Machine Learning · Computer Science 2025-05-27 Siqi Kou , Qingyuan Tian , Hanwen Xu , Zihao Zeng , Zhijie Deng

In-context learning is a promising paradigm that utilizes in-context examples as prompts for the predictions of large language models. These prompts are crucial for achieving strong performance. However, since the prompts need to be sampled…

Computation and Language · Computer Science 2025-07-15 Shaokun Zhang , Xiaobo Xia , Zhaoqing Wang , Ling-Hao Chen , Jiale Liu , Qingyun Wu , Tongliang Liu

Data attribution seeks to trace model behavior back to the training examples that shaped it, enabling debugging, auditing, and data valuation at scale. Classical influence-function methods offer a principled foundation but remain…

Machine Learning · Computer Science 2025-11-26 Sibo Ma , Julian Nyarko

Off-policy evaluation in reinforcement learning offers the chance of using observational data to improve future outcomes in domains such as healthcare and education, but safe deployment in high stakes settings requires ways of assessing its…

Machine Learning · Computer Science 2020-08-12 Omer Gottesman , Joseph Futoma , Yao Liu , Sonali Parbhoo , Leo Anthony Celi , Emma Brunskill , Finale Doshi-Velez

In reinforcement learning (RL) with experience replay, experiences stored in a replay buffer influence the RL agent's performance. Information about how these experiences influence the agent's performance is valuable for various purposes,…

Machine Learning · Computer Science 2025-07-22 Takuya Hiraoka , Guanquan Wang , Takashi Onishi , Yoshimasa Tsuruoka

In-context Learning (ICL) is the ability of Large Language Models (LLMs) to perform new tasks when conditioned on prompts comprising a few task examples. However, ICL performance can be critically sensitive to the choice of examples. To…

Computation and Language · Computer Science 2024-02-23 Shivanshu Gupta , Clemens Rosenbaum , Ethan R. Elenberg

Addressing data integrity challenges, such as unlearning the effects of data poisoning after model training, is necessary for the reliable deployment of machine learning models. State-of-the-art influence functions, such as EK-FAC and TRAK,…

Computer Vision and Pattern Recognition · Computer Science 2025-10-21 Wenjie Li , Jiawei Li , Pengcheng Zeng , Christian Schroeder de Witt , Ameya Prabhu , Amartya Sanyal