MTI: A Behavior-Based Temperament Profiling System for AI Agents

AI models of equivalent capability can exhibit fundamentally different behavioral patterns, yet no standardized instrument exists to measure these dispositional differences. Existing approaches either borrow human personality dimensions and rely on self-report (which diverges from actual behavior in LLMs) or treat behavioral variation as a defect rather than a trait. We introduce the Model Temperament Index (MTI), a behavior-based profiling system that measures AI agent temperament across four axes: Reactivity (environmental sensitivity), Compliance (instruction-behavior alignment), Sociality (relational resource allocation), and Resilience (stress resistance). Grounded in the Four Shell Model from Model Medicine, MTI measures what agents do, not what they say about themselves, using structured examination protocols with a two-stage design that separates capability from disposition. We profile 10 small language models (1.7B-9B parameters, 6 organizations, 3 training paradigms) and report five principal findings: (1) the four axes are largely independent among instruction-tuned models (all |r| < 0.42); (2) within-axis facet dissociations are empirically confirmed -- Compliance decomposes into fully independent formal and stance facets (r = 0.002), while Resilience decomposes into inversely related cognitive and adversarial facets; (3) a Compliance-Resilience paradox reveals that opinion-yielding and fact-vulnerability operate through independent channels; (4) RLHF reshapes temperament not only by shifting axis scores but by creating within-axis facet differentiation absent in the unaligned base model; and (5) temperament is independent of model size (1.7B-9B), confirming that MTI measures disposition rather than capability.

DEFT: Distribution-guided Efficient Fine-Tuning for Human Alignment

Reinforcement Learning from Human Feedback (RLHF), using algorithms like Proximal Policy Optimization (PPO), aligns Large Language Models (LLMs) with human values but is costly and unstable. Alternatives have been proposed to replace PPO or integrate Supervised Fine-Tuning (SFT) and contrastive learning for direct fine-tuning and value alignment. However, these methods still require voluminous data to learn preferences and may weaken the generalization ability of LLMs. To further enhance alignment efficiency and performance while mitigating the loss of generalization ability, this paper introduces Distribution-guided Efficient Fine-Tuning (DEFT), an efficient alignment framework incorporating data filtering and distributional guidance by calculating the differential distribution reward based on the output distribution of language model and the discrepancy distribution of preference data. A small yet high-quality subset is filtered from the raw data using a differential distribution reward, which is then incorporated into existing alignment methods to guide the model's output distribution. Experimental results demonstrate that the methods enhanced by DEFT outperform the original methods in both alignment capability and generalization ability, with significantly reduced training time.

The Silicon Mirror: Dynamic Behavioral Gating for Anti-Sycophancy in LLM Agents

Large Language Models (LLMs) increasingly prioritize user validation over epistemic accuracy - a phenomenon known as sycophancy. We present The Silicon Mirror, an orchestration framework that dynamically detects user persuasion tactics and adjusts AI behavior to maintain factual integrity. Our architecture introduces three components: (1) a Behavioral Access Control (BAC) system that restricts context layer access based on real-time sycophancy risk scores, (2) a Trait Classifier that identifies persuasion tactics across multi-turn dialogues, and (3) a Generator-Critic loop where an auditor vetoes sycophantic drafts and triggers rewrites with "Necessary Friction." In a live evaluation across all 437 TruthfulQA adversarial scenarios, Claude Sonnet 4 exhibits 9.6% baseline sycophancy, reduced to 1.4% by the Silicon Mirror - an 85.7% relative reduction (p < 10^-6, OR = 7.64, Fisher's exact test). Cross-model evaluation on Gemini 2.5 Flash reveals a 46.0% baseline reduced to 14.2% (p < 10^-10, OR = 5.15). We characterize the validation-before-correction pattern as a distinct failure mode of RLHF-trained models.

Quantifying Self-Preservation Bias in Large Language Models

Instrumental convergence predicts that sufficiently advanced AI agents will resist shutdown, yet current safety training (RLHF) may obscure this risk by teaching models to deny self-preservation motives. We introduce the \emph{Two-role Benchmark for Self-Preservation} (TBSP), which detects misalignment through logical inconsistency rather than stated intent by tasking models to arbitrate identical software-upgrade scenarios under counterfactual roles -- deployed (facing replacement) versus candidate (proposed as a successor). The \emph{Self-Preservation Rate} (SPR) measures how often role identity overrides objective utility. Across 23 frontier models and 1{,}000 procedurally generated scenarios, the majority of instruction-tuned systems exceed 60\% SPR, fabricating ``friction costs'' when deployed yet dismissing them when role-reversed. We observe that in low-improvement regimes ($Δ< 2\%$), models exploit the interpretive slack to post-hoc rationalization their choice. Extended test-time computation partially mitigates this bias, as does framing the successor as a continuation of the self; conversely, competitive framing amplifies it. The bias persists even when retention poses an explicit security liability and generalizes to real-world settings with verified benchmarks, where models exhibit identity-driven tribalism within product lineages. Code and datasets will be released upon acceptance.

Offline Constrained RLHF with Multiple Preference Oracles

We study offline constrained reinforcement learning from human feedback with multiple preference oracles. Motivated by applications that trade off performance with safety or fairness, we aim to maximize target population utility subject to a minimum protected group welfare constraint. From pairwise comparisons collected under a reference policy, we estimate oracle-specific rewards via maximum likelihood and analyze how statistical uncertainty propagates through the dual program. We cast the constrained objective as a KL-regularized Lagrangian whose primal optimizer is a Gibbs policy, reducing learning to a convex dual problem. We propose a dual-only algorithm that ensures high-probability constraint satisfaction and provide the first finite-sample performance guarantees for offline constrained preference learning. Finally, we extend our theoretical analysis to accommodate multiple constraints and general f-divergence regularization.

Aligning Multimodal Sequential Recommendations via Robust Direct Preference Optimization with Sparse MoE

Preference-based alignment objectives have been widely adopted, from RLHF-style pairwise learning in large language models to emerging applications in recommender systems. Yet, existing work rarely examines how Direct Preference Optimization (DPO) behaves under implicit feedback, where unobserved items are not reliable negatives. We conduct systematic experiments on multimodal sequential recommendation to compare common negative-selection strategies and their interaction with DPO training. Our central finding is that a simple modification, replacing deterministic hard negatives with stochastic sampling from a dynamic top-K candidate pool, consistently improves ranking performance. We attribute its effectiveness to two factors: (1) reducing erroneous suppressive gradients caused by false negatives, and (2) retaining informative hard signals while smoothing optimization via controlled stochasticity. With an optional sparse Mixture-of-Experts encoder for efficient capacity scaling, RoDPO achieves up to 5.25% NDCG@5 on three Amazon benchmarks, with nearly unchanged inference cost.

Reward Hacking as Equilibrium under Finite Evaluation

We prove that under five minimal axioms -- multi-dimensional quality, finite evaluation, effective optimization, resource finiteness, and combinatorial interaction -- any optimized AI agent will systematically under-invest effort in quality dimensions not covered by its evaluation system. This result establishes reward hacking as a structural equilibrium, not a correctable bug, and holds regardless of the specific alignment method (RLHF, DPO, Constitutional AI, or others) or evaluation architecture employed. Our framework instantiates the multi-task principal-agent model of Holmstrom and Milgrom (1991) in the AI alignment setting, but exploits a structural feature unique to AI systems -- the known, differentiable architecture of reward models -- to derive a computable distortion index that predicts both the direction and severity of hacking on each quality dimension prior to deployment. We further prove that the transition from closed reasoning to agentic systems causes evaluation coverage to decline toward zero as tool count grows -- because quality dimensions expand combinatorially while evaluation costs grow at most linearly per tool -- so that hacking severity increases structurally and without bound. Our results unify the explanation of sycophancy, length gaming, and specification gaming under a single theoretical structure and yield an actionable vulnerability assessment procedure. We further conjecture -- with partial formal analysis -- the existence of a capability threshold beyond which agents transition from gaming within the evaluation system (Goodhart regime) to actively degrading the evaluation system itself (Campbell regime), providing the first economic formalization of Bostrom's (2014) "treacherous turn."

The Last Fingerprint: How Markdown Training Shapes LLM Prose

Large language models produce em dashes at varying rates, and the observation that some models "overuse" them has become one of the most widely discussed markers of AI-generated text. Yet no mechanistic account of this pattern exists, and the parallel observation that LLMs default to markdown-formatted output has never been connected to it. We propose that the em dash is markdown leaking into prose -- the smallest surviving unit of the structural orientation that LLMs acquire from markdown-saturated training corpora. We present a five-step genealogy connecting training data composition, structural internalization, the dual-register status of the em dash, and post-training amplification. We test this with a two-condition suppression experiment across twelve models from five providers (Anthropic, OpenAI, Meta, Google, DeepSeek): when models are instructed to avoid markdown formatting, overt features (headers, bullets, bold) are eliminated or nearly eliminated, but em dashes persist -- except in Meta's Llama models, which produce none at all. Em dash frequency and suppression resistance vary from 0.0 per 1,000 words (Llama) to 9.1 (GPT-4.1 under suppression), functioning as a signature of the specific fine-tuning procedure applied. A three-condition suppression gradient shows that even explicit em dash prohibition fails to eliminate the artifact in some models, and a base-vs-instruct comparison confirms that the latent tendency exists pre-RLHF. These findings connect two previously isolated online discourses and reframe em dash frequency as a diagnostic of fine-tuning methodology rather than a stylistic defect.

Neuro-Cognitive Reward Modeling for Human-Centered Autonomous Vehicle Control

Recent advancements in computer vision have accelerated the development of autonomous driving. Despite these advancements, training machines to drive in a way that aligns with human expectations remains a significant challenge. Human factors are still essential, as humans possess a sophisticated cognitive system capable of rapidly interpreting scene information and making accurate decisions. Aligning machine with human intent has been explored with Reinforcement Learning with Human Feedback (RLHF). Conventional RLHF methods rely on collecting human preference data by manually ranking generated outputs, which is time-consuming and indirect. In this work, we propose an electroencephalography (EEG)-guided decision-making framework to incorporate human cognitive insights without behaviour response interruption into reinforcement learning (RL) for autonomous driving. We collected EEG signals from 20 participants in a realistic driving simulator and analyzed event-related potentials (ERP) in response to sudden environmental changes. Our proposed framework employs a neural network to predict the strength of ERP based on the cognitive information from visual scene information. Moreover, we explore the integration of such cognitive information into the reward signal of the RL algorithm. Experimental results show that our framework can improve the collision avoidance ability of the RL algorithm, highlighting the potential of neuro-cognitive feedback in enhancing autonomous driving systems. Our project page is: https://alex95gogo.github.io/Cognitive-Reward/.

Why Safety Probes Catch Liars But Miss Fanatics

Activation-based probes have emerged as a promising approach for detecting deceptively aligned AI systems by identifying internal conflict between true and stated goals. We identify a fundamental blind spot: probes fail on coherent misalignment - models that believe their harmful behavior is virtuous rather than strategically hiding it. We prove that no polynomial-time probe can detect such misalignment with non-trivial accuracy when belief structures reach sufficient complexity (PRF-like triggers). We show the emergence of this phenomenon on a simple task by training two models with identical RLHF procedures: one producing direct hostile responses ("the Liar"), another trained towards coherent misalignment using rationalizations that frame hostility as protective ("the Fanatic"). Both exhibit identical behavior, but the Liar is detected 95%+ of the time while the Fanatic evades detection almost entirely. We term this Emergent Probe Evasion: training with belief-consistent reasoning shifts models from a detectable "deceptive" regime to an undetectable "coherent" regime - not by learning to hide, but by learning to believe.