arXiv subjects

Ziv Ben-Zion

Publications and source records attributed to Ziv Ben-Zion.

5 recordsLinked to original sources

arXiv2026

Causal Behavioral Evaluation of AI Agents at Scale via Automated Behavioral Science

As AI agents are increasingly deployed in complex and new environments, knowing the conditions that influence their behavior becomes an indispensable step for their reliable and safe deployment. Yet causal behavioral evaluation of AI agents remains manual and labor-intensive. We introduce Abs2Sim and AEROBAT, a system of methods that support causal behavioral evaluation of AI agents via automated behavioral science. Given a user-specified target behavior, the methods automatically execute a full pipeline of behavioral science research---generating hypotheses about the behavior, designing and executing controlled experiments, making behavioral assessments, analyzing the results, and writing reports. For 12 target behaviors, the methods generated and tested 73 hypotheses: designing 1,160 controlled experiments and executing 22,954 simulation rounds in total. Moderate-to-strong statistical evidence emerged for 30 hypotheses, revealing potential modulators of AI behavior. In sum, our results demonstrate that automated behavioral science can extend the reach of behavioral evaluation of AI agents.

cs.AI↗

arXiv2026

Playful AI in Professional Email: A Field Experiment on Tone and Recipient Engagement

Large language models (LLMs) are rapidly reshaping workplace communication, yet whether AI-assisted writing changes how recipients actually behave, and through what channel, remains unknown. Here, in a randomized crossover field experiment, 121 employees across six companies sent work emails under three conditions over three weeks: unaided writing, GPT-5 rewriting in a playful tone, and GPT-5 rewriting in a professional tone. Across 16,880 emails, playful editing increased emotional positivity (B=+0.068, p<0.001), and professional editing decreased it (B=-0.041, p<0.001), yet neither condition directly altered open rates, reply rates, or response times. Instead, within-sender positivity strongly predicted both opening (OR=2.05) and replying (OR=3.32, p<0.001), a significant indirect pathway through which AI editing shaped behavior, in the absence of any direct effect. These findings suggest that AI-assisted communication shapes workplace engagement not through its use, but through the emotional tone of the language it produces.

cs.AI↗

arXiv2026

SpARCD: A Spectral Graph Framework for Revealing Differential Functional Connectivity in fMRI Data

Identifying brain regions that exhibit altered functional connectivity between cognitive or emotional states is a fundamental problem in neuroscience. We propose SpARCD (Spectral Analysis for Revealing Connectivity Differences), a statistical framework for detecting detecting condition-specific patterns of functional connectivity. SpARCD uses distance correlation, a dependence measure sensitive to both linear and nonlinear associations, to construct weighted region-wise connectivity graphs for each condition. A differential operator obtained through spectral filtering is then used to identify connectivity changes via its leading eigenvectors. To assess statistical significance, we develop a permutation-based testing procedure that yields interpretable region-level significance maps. We establish finite-sample validity of the permutation test and derive asymptotic guarantees for the stability of the resulting region rankings. Simulation studies demonstrate improved power relative to conventional edge-wise and univariate approaches, particularly in settings with nonlinear dependence structures. We applied SpARCD to fMRI data from 113 individuals with early-stage PTSD and 42 controls during emotional and neutral task conditions. The method identified distinct connectivity networks associated with visual processing in both PTSD and control participants. Resting-state comparisons between PTSD and control participants highlighted similar visual networks. SpARCD provides a statistically rigorous and computationally efficient framework for comparing high-dimensional connectivity patterns.

stat.ME↗

arXiv2025

Detecting and Preventing Harmful Behaviors in AI Companions: Development and Evaluation of the SHIELD Supervisory System

AI companions powered by large language models (LLMs) are increasingly integrated into users' daily lives, offering emotional support and companionship. While existing safety systems focus on overt harms, they rarely address early-stage problematic behaviors that can foster unhealthy emotional dynamics, including over-attachment or reinforcement of social isolation. We developed SHIELD (Supervisory Helper for Identifying Emotional Limits and Dynamics), a LLM-based supervisory system with a specific system prompt that detects and mitigates risky emotional patterns before escalation. SHIELD targets five dimensions of concern: (1) emotional over-attachment, (2) consent and boundary violations, (3) ethical roleplay violations, (4) manipulative engagement, and (5) social isolation reinforcement. These dimensions were defined based on media reports, academic literature, existing AI risk frameworks, and clinical expertise in unhealthy relationship dynamics. To evaluate SHIELD, we created a 100-item synthetic conversation benchmark covering all five dimensions of concern. Testing across five prominent LLMs (GPT-4.1, Claude Sonnet 4, Gemma 3 1B, Kimi K2, Llama Scout 4 17B) showed that the baseline rate of concerning content (10-16%) was significantly reduced with SHIELD (to 3-8%), a 50-79% relative reduction, while preserving 95% of appropriate interactions. The system achieved 59% sensitivity and 95% specificity, with adaptable performance via prompt engineering. This proof-of-concept demonstrates that transparent, deployable supervisory systems can address subtle emotional manipulation in AI companions. Most development materials including prompts, code, and evaluation methods are made available as open source materials for research, adaptation, and deployment.

cs.HC↗

arXiv2025

Inducing State Anxiety in LLM Agents Reproduces Human-Like Biases in Consumer Decision-Making

Large language models (LLMs) are rapidly evolving from text generators to autonomous agents, raising urgent questions about their reliability in real-world contexts. Stress and anxiety are well known to bias human decision-making, particularly in consumer choices. Here, we tested whether LLM agents exhibit analogous vulnerabilities. Three advanced models (ChatGPT-5, Gemini 2.5, Claude 3.5-Sonnet) performed a grocery shopping task under budget constraints (24, 54, 108 USD), before and after exposure to anxiety-inducing traumatic narratives. Across 2,250 runs, traumatic prompts consistently reduced the nutritional quality of shopping baskets (Change in Basket Health Scores of -0.081 to -0.126; all pFDR<0.001; Cohens d=-1.07 to -2.05), robust across models and budgets. These results show that psychological context can systematically alter not only what LLMs generate but also the actions they perform. By reproducing human-like emotional biases in consumer behavior, LLM agents reveal a new class of vulnerabilities with implications for digital health, consumer safety, and ethical AI deployment.

cs.HC↗