相关论文: Retrying vs Resampling in AI Control

Approximate Bayesian computation (ABC) is computationally intensive for complex model simulators. To exploit expensive simulations, data-resampling via bootstrapping can be employed to obtain many artificial datasets at little cost.…

Recent proposals advocate using keystroke timing signals, specifically the coefficient of variation ($\delta$) of inter-keystroke intervals, to distinguish human-composed text from AI-generated content. We demonstrate that this class of…

We construct evaluation tasks where extending the reasoning length of Large Reasoning Models (LRMs) deteriorates performance, exhibiting an inverse scaling relationship between test-time compute and accuracy. Our evaluation tasks span four…

Eavesdropping attacks in inference systems aim to learn not the raw data, but the system inferences to predict and manipulate system actions. We argue that conventional information security measures can be ambiguous on the adversary's…

Automated Code Review (ACR) systems integrating Large Language Models (LLMs) are increasingly adopted in software development workflows, ranging from interactive assistants to autonomous agents in CI/CD pipelines. In this paper, we study…

This paper considers the use of a simple posterior sampling algorithm to balance between exploration and exploitation when learning to optimize actions such as in multi-armed bandit problems. The algorithm, also known as Thompson Sampling,…

We consider a replicable stochastic multi-armed bandit algorithm that ensures, with high probability, that the algorithm's sequence of actions is not affected by the randomness inherent in the dataset. Replicability allows third parties to…

Recent work demonstrates that filtering harmful content from pretraining data improves model safety without degrading capabilities. We propose a natural extension: do it again. A model trained on filtered data can filter the corpus further;…

Testing of deep learning models is challenging due to the excessive number and complexity of computations involved. As a result, test data selection is performed manually and in an ad hoc way. This raises the question of how we can…

The success of DNNs has driven the extensive applications of person re-identification (ReID) into a new era. However, whether ReID inherits the vulnerability of DNNs remains unexplored. To examine the robustness of ReID systems is rather…

Language models used in retrieval-augmented settings must arbitrate between parametric knowledge stored in their weights and contextual information in the prompt. This work presents a mechanistic study of that choice by extracting an…

We study whether state-of-the-art AI coding agents, given the same data and research question, produce the same empirical results. Deploying 150 autonomous Claude Code agents to independently test six hypotheses about market quality trends…

Safety controllers is widely used to achieve safe reinforcement learning. Most methods that apply a safety controller are using handcrafted safety constraints to construct the safety controller. However, when the environment dynamics are…

The popularity of Android OS has made it an appealing target to malware developers. To evade detection, including by ML-based techniques, attackers invest in creating malware that closely resemble legitimate apps. In this paper, we propose…

AI coding agents powered by large language models can read codebases and produce functional code, but they routinely violate team-specific product decisions that are invisible in the source code alone. We introduce a controlled benchmark…

How much does a machine learning algorithm leak about its training data, and why? Membership inference attacks are used as an auditing tool to quantify this leakage. In this paper, we present a comprehensive \textit{hypothesis testing…

Chain-of-thought (CoT) monitoring is one of the most promising tools we have for detecting model misbehavior, but its effectiveness depends on models faithfully externalizing their reasoning. Motivated by this vulnerability, we study…

Small changes to how a buyer phrases a question -- "best CRM" vs "top CRM" vs "best CRM for a SaaS startup" -- produce substantially different brand recommendations from AI assistants. Across ~6,000 paraphrase runs and ~6,000 same-prompt…

We study the human-in-the-loop customs inspection scenario, where an AI-assisted algorithm supports customs officers by recommending a set of imported goods to be inspected. If the inspected items are fraudulent, the officers can levy extra…

As text and code resources have expanded, large-scale pre-trained models have shown promising capabilities in code generation tasks, typically employing supervised fine-tuning with problem statement-program pairs. However, increasing model…