Teaching AI to Handle Exceptions: Supervised Fine-Tuning with Human-Aligned Judgment

• URL: http://arxiv.org/abs/2503.02976v1
• Date: Tue, 04 Mar 2025 20:00:37 GMT
• Title: Teaching AI to Handle Exceptions: Supervised Fine-Tuning with Human-Aligned Judgment
• Authors: Matthew DosSantos DiSorbo, Harang Ju, Sinan Aral,
• Abstract summary: Large language models (LLMs) are evolving into agentic AI systems, but their decision-making processes remain poorly understood.<n>We show that even LLMs that excel at reasoning deviate significantly from human judgments because they adhere strictly to policies.<n>We then evaluate three approaches to tuning AI agents to handle exceptions: ethical framework prompting, chain-of-thought reasoning, and supervised fine-tuning.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs), initially developed for generative AI, are now evolving into agentic AI systems, which make decisions in complex, real-world contexts. Unfortunately, while their generative capabilities are well-documented, their decision-making processes remain poorly understood. This is particularly evident when models are handling exceptions, a critical and challenging aspect of decision-making made relevant by the inherent incompleteness of contracts. Here we demonstrate that LLMs, even ones that excel at reasoning, deviate significantly from human judgments because they adhere strictly to policies, even when such adherence is impractical, suboptimal, or even counterproductive. We then evaluate three approaches to tuning AI agents to handle exceptions: ethical framework prompting, chain-of-thought reasoning, and supervised fine-tuning. We find that while ethical framework prompting fails and chain-of-thought prompting provides only slight improvements, supervised fine-tuning, specifically with human explanations, yields markedly better results. Surprisingly, in our experiments, supervised fine-tuning even enabled models to generalize human-like decision-making to novel scenarios, demonstrating transfer learning of human-aligned decision-making across contexts. Furthermore, fine-tuning with explanations, not just labels, was critical for alignment, suggesting that aligning LLMs with human judgment requires explicit training on how decisions are made, not just which decisions are made. These findings highlight the need to address LLMs' shortcomings in handling exceptions in order to guide the development of agentic AI toward models that can effectively align with human judgment and simultaneously adapt to novel contexts.

Related papers

• Do LLMs trust AI regulation? Emerging behaviour of game-theoretic LLM agents [61.132523071109354]
  This paper investigates the interplay between AI developers, regulators and users, modelling their strategic choices under different regulatory scenarios. Our research identifies emerging behaviours of strategic AI agents, which tend to adopt more "pessimistic" stances than pure game-theoretic agents.
  arXiv  Detail & Related papers  (2025-04-11T15:41:21Z)
• MR-Ben: A Meta-Reasoning Benchmark for Evaluating System-2 Thinking in LLMs [55.20845457594977]
  Large language models (LLMs) have shown increasing capability in problem-solving and decision-making.<n>We present a process-based benchmark MR-Ben that demands a meta-reasoning skill.<n>Our meta-reasoning paradigm is especially suited for system-2 slow thinking.
  arXiv  Detail & Related papers  (2024-06-20T03:50:23Z)
• Aligning Large Language Models from Self-Reference AI Feedback with one General Principle [61.105703857868775]
  We propose a self-reference-based AI feedback framework that enables a 13B Llama2-Chat to provide high-quality feedback. Specifically, we allow the AI to first respond to the user's instructions, then generate criticism of other answers based on its own response as a reference. Finally, we determine which answer better fits human preferences according to the criticism.
  arXiv  Detail & Related papers  (2024-06-17T03:51:46Z)
• Tuning-Free Accountable Intervention for LLM Deployment -- A Metacognitive Approach [55.613461060997004]
  Large Language Models (LLMs) have catalyzed transformative advances across a spectrum of natural language processing tasks. We propose an innovative textitmetacognitive approach, dubbed textbfCLEAR, to equip LLMs with capabilities for self-aware error identification and correction.
  arXiv  Detail & Related papers  (2024-03-08T19:18:53Z)
• On the meaning of uncertainty for ethical AI: philosophy and practice [10.591284030838146]
  We argue that this is a significant way to bring ethical considerations into mathematical reasoning. We demonstrate these ideas within the context of competing models used to advise the UK government on the spread of the Omicron variant of COVID-19 during December 2021.
  arXiv  Detail & Related papers  (2023-09-11T15:13:36Z)
• In Search of Verifiability: Explanations Rarely Enable Complementary Performance in AI-Advised Decision Making [25.18203172421461]
  We argue explanations are only useful to the extent that they allow a human decision maker to verify the correctness of an AI's prediction. We also compare the objective of complementary performance with that of appropriate reliance, decomposing the latter into the notions of outcome-graded and strategy-graded reliance.
  arXiv  Detail & Related papers  (2023-05-12T18:28:04Z)
• Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervision [84.31474052176343]
  Recent AI-assistant agents, such as ChatGPT, rely on supervised fine-tuning (SFT) with human annotations and reinforcement learning from human feedback to align the output with human intentions. This dependence can significantly constrain the true potential of AI-assistant agents due to the high cost of obtaining human supervision. We propose a novel approach called SELF-ALIGN, which combines principle-driven reasoning and the generative power of LLMs for the self-alignment of AI agents with minimal human supervision.
  arXiv  Detail & Related papers  (2023-05-04T17:59:28Z)
• Inverse Online Learning: Understanding Non-Stationary and Reactionary Policies [79.60322329952453]
  We show how to develop interpretable representations of how agents make decisions. By understanding the decision-making processes underlying a set of observed trajectories, we cast the policy inference problem as the inverse to this online learning problem. We introduce a practical algorithm for retrospectively estimating such perceived effects, alongside the process through which agents update them. Through application to the analysis of UNOS organ donation acceptance decisions, we demonstrate that our approach can bring valuable insights into the factors that govern decision processes and how they change over time.
  arXiv  Detail & Related papers  (2022-03-14T17:40:42Z)
• Learning Behavioral Soft Constraints from Demonstrations [31.34800444313487]
  We present a novel inverse reinforcement learning (IRL) method for learning implicit hard and soft constraints over states, actions, and state features. Our method enables agents implicitly learn human constraints and desires without the need for explicit modeling by the agent designer.
  arXiv  Detail & Related papers  (2022-02-21T18:09:56Z)
• Making Human-Like Trade-offs in Constrained Environments by Learning from Demonstrations [30.738257457765755]
  We present a novel inverse reinforcement learning (IRL) method for learning implicit hard and soft constraints from demonstrations. We then use the constraint learning method to implement a novel system architecture that orchestrates competing objectives. We evaluate the resulting agent on trajectory length, number of violated constraints, and total reward, demonstrating that our agent architecture is both general and achieves strong performance.
  arXiv  Detail & Related papers  (2021-09-22T20:12:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.