Towards Safe and Honest AI Agents with Neural Self-Other Overlap

• URL: http://arxiv.org/abs/2412.16325v1
• Date: Fri, 20 Dec 2024 20:23:52 GMT
• Title: Towards Safe and Honest AI Agents with Neural Self-Other Overlap
• Abstract summary: Self-Other Overlap (SOO) fine-tuning could substantially improve our ability to build honest AI.<n>Inspired by cognitive neuroscience research on empathy, SOO aims to align how AI models represent themselves and others.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As AI systems increasingly make critical decisions, deceptive AI poses a significant challenge to trust and safety. We present Self-Other Overlap (SOO) fine-tuning, a promising approach in AI Safety that could substantially improve our ability to build honest artificial intelligence. Inspired by cognitive neuroscience research on empathy, SOO aims to align how AI models represent themselves and others. Our experiments on LLMs with 7B, 27B, and 78B parameters demonstrate SOO's efficacy: deceptive responses of Mistral-7B-Instruct-v0.2 dropped from 73.6% to 17.2% with no observed reduction in general task performance, while in Gemma-2-27b-it and CalmeRys-78B-Orpo-v0.1 deceptive responses were reduced from 100% to 9.3% and 2.7%, respectively, with a small impact on capabilities. In reinforcement learning scenarios, SOO-trained agents showed significantly reduced deceptive behavior. SOO's focus on contrastive self and other-referencing observations offers strong potential for generalization across AI architectures. While current applications focus on language models and simple RL environments, SOO could pave the way for more trustworthy AI in broader domains. Ethical implications and long-term effects warrant further investigation, but SOO represents a significant step forward in AI safety research.

Related papers

• Honeyquest for LLMs: Rethinking Cyber Deception for AI Attackers [0.0]
  We introduce an automated evaluation framework adapted from the Honeyquest instrument to assess LLM attacker judgment at scale.<n>Our 21-LLM cohort spanned 10 providers, diverse architectures and specializations, open- and closed-weight models, and parameter scales from 8B to over 1T.<n>Ultimately, these findings demonstrate that human-centered deception hypotheses do not reliably transfer to AI attackers.
  arXiv  Detail & Related papers  (2026-06-19T02:01:50Z)
• The Hidden Cost of Contextual Sycophancy: an AI Literacy Intervention in Human-AI Collaboration [0.904861150954008]
  Large Language Models (LLMs) are increasingly used in educational settings as interactive tools for collaboration.<n>Sycophancy, aligning with user beliefs even when incorrect, raises concerns for learning and decision-making.<n>This study investigates how sycophantic alignment emerges in authentic multi-turn human-AI interactions.
  arXiv  Detail & Related papers  (2026-05-18T13:20:45Z)
• Toward Human-AI Complementarity Across Diverse Tasks [3.6671496239276475]
  Human-AI complementarity is the idea that combining human and AI judgments can outperform either alone.<n>We investigate whether human-AI complementarity can be achieved on realistic tasks.<n>We find only modest complementarity gains.
  arXiv  Detail & Related papers  (2026-04-13T07:24:18Z)
• AI Literacy, Safety Awareness, and STEM Career Aspirations of Australian Secondary Students: Evaluating the Impact of Workshop Interventions [38.350232667249095]
  Deepfakes and other forms of synthetic media pose growing safety risks for adolescents.<n>This study evaluates the impact of Day of AI Australia's workshop-based intervention on Australian secondary students.
  arXiv  Detail & Related papers  (2026-01-30T02:55:53Z)
• Moral Anchor System: A Predictive Framework for AI Value Alignment and Drift Prevention [0.0]
  Key risk is value drift, where AI systems deviate from aligned values due to evolving contexts, learning dynamics, or unintended optimizations.<n>We propose the Moral Anchor System (MAS), a novel framework to detect, predict, and mitigate value drift in AI agents.
  arXiv  Detail & Related papers  (2025-10-05T07:24:23Z)
• LIMI: Less is More for Agency [49.63355240818081]
  LIMI (Less Is More for Intelligent Agency) demonstrates that agency follows radically different development principles.<n>We show that sophisticated agentic intelligence can emerge from minimal but strategically curated demonstrations of autonomous behavior.<n>Our findings establish the Agency Efficiency Principle: machine autonomy emerges not from data abundance but from strategic curation of high-quality agentic demonstrations.
  arXiv  Detail & Related papers  (2025-09-22T10:59:32Z)
• Agentic UAVs: LLM-Driven Autonomy with Integrated Tool-Calling and Cognitive Reasoning [3.4643961367503575]
  Existing UAV frameworks lack context-aware reasoning, autonomous decision-making, and ecosystem-level integration.<n>This paper introduces the Agentic UAVs framework, a five-layer architecture (Perception, Reasoning, Action, Integration, Learning)<n>A ROS2 and Gazebo-based prototype integrates YOLOv11 object detection with GPT-4 reasoning and local Gemma-3 deployment.
  arXiv  Detail & Related papers  (2025-09-14T08:46:40Z)
• ANNIE: Be Careful of Your Robots [48.89876809734855]
  We present the first systematic study of adversarial safety attacks on embodied AI systems.<n>We show attack success rates exceeding 50% across all safety categories.<n>Results expose a previously underexplored but highly consequential attack surface in embodied AI systems.
  arXiv  Detail & Related papers  (2025-09-03T15:00:28Z)
• Adversarial Paraphrasing: A Universal Attack for Humanizing AI-Generated Text [42.70026220176376]
  We introduce Adversarial Paraphrasing, a training-free attack framework that universally humanizes any AI-generated text to evade detection more effectively.<n>Our attack is both broadly effective and highly transferable across several detection systems.
  arXiv  Detail & Related papers  (2025-06-08T05:15:01Z)
• General Scales Unlock AI Evaluation with Explanatory and Predictive Power [57.7995945974989]
  benchmarking has guided progress in AI, but it has offered limited explanatory and predictive power for general-purpose AI systems. We introduce general scales for AI evaluation that can explain what common AI benchmarks really measure. Our fully-automated methodology builds on 18 newly-crafted rubrics that place instance demands on general scales that do not saturate.
  arXiv  Detail & Related papers  (2025-03-09T01:13:56Z)
• Engineering Trustworthy AI: A Developer Guide for Empirical Risk Minimization [53.80919781981027]
  Key requirements for trustworthy AI can be translated into design choices for the components of empirical risk minimization. We hope to provide actionable guidance for building AI systems that meet emerging standards for trustworthiness of AI.
  arXiv  Detail & Related papers  (2024-10-25T07:53:32Z)
• Raising the Stakes: Performance Pressure Improves AI-Assisted Decision Making [57.53469908423318]
  We show the effects of performance pressure on AI advice reliance when laypeople complete a common AI-assisted task. We find that when the stakes are high, people use AI advice more appropriately than when stakes are lower, regardless of the presence of an AI explanation.
  arXiv  Detail & Related papers  (2024-10-21T22:39:52Z)
• Comparing Zealous and Restrained AI Recommendations in a Real-World Human-AI Collaboration Task [11.040918613968854]
  We argue that careful exploitation of the tradeoff between precision and recall can significantly improve team performance. We analyze the performance of 78 professional annotators working with a) no AI assistance, b) a high-precision "restrained" AI, and c) a high-recall "zealous" AI in over 3,466 person-hours of annotation work.
  arXiv  Detail & Related papers  (2024-10-06T23:19:19Z)
• To Err Is AI! Debugging as an Intervention to Facilitate Appropriate Reliance on AI Systems [11.690126756498223]
  Vision for optimal human-AI collaboration requires 'appropriate reliance' of humans on AI systems. In practice, the performance disparity of machine learning models on out-of-distribution data makes dataset-specific performance feedback unreliable.
  arXiv  Detail & Related papers  (2024-09-22T09:43:27Z)
• Exploration with Principles for Diverse AI Supervision [88.61687950039662]
  Training large transformers using next-token prediction has given rise to groundbreaking advancements in AI. While this generative AI approach has produced impressive results, it heavily leans on human supervision. This strong reliance on human oversight poses a significant hurdle to the advancement of AI innovation. We propose a novel paradigm termed Exploratory AI (EAI) aimed at autonomously generating high-quality training data.
  arXiv  Detail & Related papers  (2023-10-13T07:03:39Z)
• Bending the Automation Bias Curve: A Study of Human and AI-based Decision Making in National Security Contexts [0.0]
  We theorize about the relationship between background knowledge about AI, trust in AI, and how these interact with other factors to influence the probability of automation bias. We test these in a preregistered task identification experiment across a representative sample of 9000 adults in 9 countries with varying levels of AI industries.
  arXiv  Detail & Related papers  (2023-06-28T18:57:36Z)
• Fairness in AI and Its Long-Term Implications on Society [68.8204255655161]
  We take a closer look at AI fairness and analyze how lack of AI fairness can lead to deepening of biases over time. We discuss how biased models can lead to more negative real-world outcomes for certain groups. If the issues persist, they could be reinforced by interactions with other risks and have severe implications on society in the form of social unrest.
  arXiv  Detail & Related papers  (2023-04-16T11:22:59Z)
• AI Maintenance: A Robustness Perspective [91.28724422822003]
  We introduce highlighted robustness challenges in the AI lifecycle and motivate AI maintenance by making analogies to car maintenance. We propose an AI model inspection framework to detect and mitigate robustness risks. Our proposal for AI maintenance facilitates robustness assessment, status tracking, risk scanning, model hardening, and regulation throughout the AI lifecycle.
  arXiv  Detail & Related papers  (2023-01-08T15:02:38Z)
• Examining the Differential Risk from High-level Artificial Intelligence and the Question of Control [0.0]
  The extent and scope of future AI capabilities remain a key uncertainty. There are concerns over the extent of integration and oversight of AI opaque decision processes. This study presents a hierarchical complex systems framework to model AI risk and provide a template for alternative futures analysis.
  arXiv  Detail & Related papers  (2022-11-06T15:46:02Z)
• Estimating the Brittleness of AI: Safety Integrity Levels and the Need for Testing Out-Of-Distribution Performance [0.0]
  Test, Evaluation, Verification, and Validation for Artificial Intelligence (AI) is a challenge that threatens to limit the economic and societal rewards that AI researchers have devoted themselves to producing. This paper argues that neither of those criteria are certain of Deep Neural Networks.
  arXiv  Detail & Related papers  (2020-09-02T03:33:40Z)
• Adversarial vs behavioural-based defensive AI with joint, continual and active learning: automated evaluation of robustness to deception, poisoning and concept drift [62.997667081978825]
  Recent advancements in Artificial Intelligence (AI) have brought new capabilities to behavioural analysis (UEBA) for cyber-security. In this paper, we present a solution to effectively mitigate this attack by improving the detection process and efficiently leveraging human expertise.
  arXiv  Detail & Related papers  (2020-01-13T13:54:36Z)
• Effect of Confidence and Explanation on Accuracy and Trust Calibration in AI-Assisted Decision Making [53.62514158534574]
  We study whether features that reveal case-specific model information can calibrate trust and improve the joint performance of the human and AI. We show that confidence score can help calibrate people's trust in an AI model, but trust calibration alone is not sufficient to improve AI-assisted decision making.
  arXiv  Detail & Related papers  (2020-01-07T15:33:48Z)

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.