English
Related papers

Related papers: Learning When to Trust in Contextual Bandits

200 papers

Safe reinforcement learning (Safe RL) aims to ensure policy performance while satisfying safety constraints. However, most existing Safe RL methods assume benign environments, making them vulnerable to adversarial perturbations commonly…

Machine Learning · Computer Science 2026-02-19 Jialiang Fan , Shixiong Jiang , Mengyu Liu , Fanxin Kong

We measure the performance of in-context learning as a function of task novelty and difficulty for open and closed questions. For that purpose, we created a novel benchmark consisting of hard scientific questions, each paired with a context…

Computation and Language · Computer Science 2024-07-03 Xiang Li , Haoran Tang , Siyu Chen , Ziwei Wang , Ryan Chen , Marcin Abram

Large Language Models (LLMs) frequently prioritize conflicting in-context information over pre-existing parametric memory, a phenomenon often termed sycophancy or compliance. However, the mechanistic realization of this behavior remains…

Machine Learning · Computer Science 2026-02-09 Long Zhang , Fangwei Lin

We introduce a stochastic contextual bandit model where at each time step the environment chooses a distribution over a context set and samples the context from this distribution. The learner observes only the context distribution while the…

Machine Learning · Statistics 2019-11-15 Johannes Kirschner , Andreas Krause

We study the problem of meta-learning several contextual stochastic bandits tasks by leveraging their concentration around a low-dimensional affine subspace, which we learn via online principal component analysis to reduce the expected…

Machine Learning · Computer Science 2024-04-02 Steven Bilaj , Sofien Dhouib , Setareh Maghsudi

With state-of-the-art models achieving high performance on standard benchmarks, contemporary research paradigms continue to emphasize general intelligence as an enduring objective. However, this pursuit overlooks the fundamental disparities…

Artificial Intelligence · Computer Science 2023-10-03 Nick DiSanto

We study sequential decision making in environments where rewards are only partially observed, but can be modeled as a function of observed contexts and the chosen action by the decision maker. This setting, known as contextual bandits,…

Methodology · Statistics 2015-03-11 Miroslav Dudík , Dumitru Erhan , John Langford , Lihong Li

Current LLM unlearning methods face a critical security vulnerability that undermines their fundamental purpose: while they appear to successfully remove sensitive or harmful knowledge, this ``forgotten" information remains precariously…

Machine Learning · Computer Science 2025-10-01 Wenhan Wu , Zheyuan Liu , Chongyang Gao , Ren Wang , Kaize Ding

Humans often become more self-aware under threat, yet can lose self-awareness when absorbed in a task; we hypothesize that language models exhibit environment-dependent \textit{evaluation awareness}. This raises concerns that models could…

Artificial Intelligence · Computer Science 2026-03-05 Maheep Chaudhary

In this work we investigate meta-learning (or learning-to-learn) approaches in multi-task linear stochastic bandit problems that can originate from multiple environments. Inspired by the work of [1] on meta-learning in a sequence of linear…

Many physical systems have underlying safety considerations that require that the strategy deployed ensures the satisfaction of a set of constraints. Further, often we have only partial information on the state of the system. We study the…

Machine Learning · Computer Science 2022-03-30 Jiabin Lin , Xian Yeow Lee , Talukder Jubery , Shana Moothedath , Soumik Sarkar , Baskar Ganapathysubramanian

While Large Language Models (LLMs) are widely documented to be sensitive to minor prompt perturbations and prone to sycophantic alignment, their robustness in consequential, rule-bound decision-making remains under-explored. We uncover a…

Artificial Intelligence · Computer Science 2026-04-07 Jon Chun , Katherine Elkins

In this paper, we consider online learning in generalized linear contextual bandits where rewards are not immediately observed. Instead, rewards are available to the decision-maker only after some delay, which is unknown and stochastic. We…

Machine Learning · Computer Science 2020-03-12 Jose Blanchet , Renyuan Xu , Zhengyuan Zhou

Large Language Models (LLMs) are often augmented with external contexts, such as those used in retrieval-augmented generation (RAG). However, these contexts can be inaccurate or intentionally misleading, leading to conflicts with the…

Computation and Language · Computer Science 2025-03-18 Yukun Huang , Sanxing Chen , Hongyi Cai , Bhuwan Dhingra

Large Language Models (LLMs) have emerged as powerful tools for diverse applications. However, their uniform token processing paradigm introduces critical vulnerabilities in instruction handling, particularly when exposed to adversarial…

Cryptography and Security · Computer Science 2025-12-04 Tengyun Ma , Jiaqi Yao , Daojing He , Shihao Peng , Yu Li , Shaohui Liu , Zhuotao Tian

Recent studies show that models trained by continual learning can achieve the comparable performances as the standard supervised learning and the learning flexibility of continual learning models enables their wide applications in the real…

Machine Learning · Computer Science 2023-04-03 Tao Bai , Chen Chen , Lingjuan Lyu , Jun Zhao , Bihan Wen

Language models (LMs) show promise for vulnerability detection but struggle with long, real-world code due to sparse and uncertain vulnerability locations. These issues, exacerbated by token limits, often cause models to miss…

Software Engineering · Computer Science 2025-07-16 Xinran Zheng , Xingzhi Qian , Huichi Zhou , Shuo Yang , Yiling He , Suman Jana , Lorenzo Cavallaro

Most off-policy evaluation methods for contextual bandits have focused on the expected outcome of a policy, which is estimated via methods that at best provide only asymptotic guarantees. However, in many applications, the expectation may…

Machine Learning · Statistics 2022-10-27 Muhammad Faaiz Taufiq , Jean-Francois Ton , Rob Cornish , Yee Whye Teh , Arnaud Doucet

Instruction-tuned large language models produce helpful, structured responses, but how robust is this helpfulness under trivial constraints? We show that simple lexical constraints (banning a single punctuation character or common word)…

Computation and Language · Computer Science 2026-04-28 Erfan Baghaei Potraghloo , Seyedarmin Azizi , Souvik Kundu , Massoud Pedram

We initiate an investigation of learning tasks in a setting where the learner is given access to two competing provers, only one of which is honest. Specifically, we consider the power of such learners in assessing purported properties of…

Machine Learning · Statistics 2026-03-13 Ran Canetti , Ephraim Linder , Connor Wagaman