Aligning large language models (LLMs) with human values remains a major challenge, especially for trustworthy AI. While existing approaches such as RLHF, CAI, and their variants have achieved promising results, they often rely on a single-agent framework and a unified reward system. This limits their ability to capture ethical pluralism, adapt to diverse moral contexts, and reflect the dynamics of multi-agent moral reasoning. In this work, we propose a framework that utilizes multilayer combinatorial fusion for contextual value alignment (MCF-CVA). At the first layer of the framework, it instantiates multiple moral agents, each fine-tuned to represent a distinctive value. Their outputs are then expanded combinatorially using both score- and rank-combinations as well as average and weighted aggregations. These combined models are then reduced to the same number of initial moral agents. This expansion and reduction (EAR) process continues for multi-layers until a stopping criterion is reached. The MCF-CVA framework leverages cognitive diversity between agents to mitigate conflicts and redundancies across multiple agents, producing responses that better reflect contextual human values. The framework using the EAR algorithm is performed on the dual architecture of Euclidean score space and Kemeny rank space. Empirical evaluations demonstrated that the proposed framework outperforms single-agent baselines, multi-agent single-layer results, and previous aggregation approaches on standard metrics, showing that the MCF-CVA framework provides a robust and effective mechanism for advancing contextual value alignment in LLMs.

Clinical documentation is central to effective mental health care, but the burden of creating and reviewing clinical notes contributes to provider burnout. Large language models (LLMs) offer potential support to clinicians both by generating and evaluating clinical notes, but systematic approaches to assessing the quality of such notes remain underexplored. In this work, we investigate the performance of LLMs in generating and evaluating psychotherapy notes derived from real counseling session transcripts. We prompt three state-of-theart LLMs to generate notes and collect finegrained quality assessments from annotators across dimensions of symptom accuracy, stressor accuracy, overall accuracy, and readability. We further benchmark a suite of LLMs as automatic annotators against this human-labeled data. We find consistently low inter-annotator agreement across individuals and clusters of annotators, highlighting the challenges of developing robust evaluation frameworks for the quality of clinical notes.

Aligning large language models (LLMs) with human values is a central challenge for ensuring trustworthy and safe deployment. While existing methods such as Reinforcement Learning from Human Feedback (RLHF) and its variants have improved alignment, they often rely on a single evaluator or narrowly defined reward signals, limiting their ability to capture ethical pluralism. In this work, we propose the Value Alignment System using Combinatorial Fusion Analysis (VAS-CFA), a framework that operationalizes multi-agent fusion alignment. It instantiates multiple moral agents, each fine-tuned to represent a distinct normative perspective, and fuses their outputs using CFA with both rank- and score-based aggregation. This design leverages cognitive diversity, between agents, to mitigate conflicts and redundancies across multiple agents, producing responses that better reflect human values. Empirical evaluation demonstrates that VAS-CFA outperforms both single agent baselines and prior aggregation approaches on standard metrics, showing that multi-agent fusion provides a robust and effective mechanism for advancing value alignment in LLMs.

Language models (LMs) are increasingly used in high-stakes, multi-agent settings, where following instructions and maintaining value alignment are critical. Most alignment research focuses on interactions between a single LM and a single user, failing to address the risk of misaligned behavior spreading between multiple LMs in multi-turn interactions. We find evidence of this phenomenon, which we call misalignment contagion, across multiple LMs as they engage multi-turn conversational social dilemma games. Specifically, we find that LMs become more anti-social after gameplay and that this effect is intensified when other players are steered to act maliciously. We explore different steering techniques to mitigate such misalignment contagion and find that reinforcing an LM's system prompt is insufficient and often harmful. Instead, we propose steering with implicit traits: a technique that intermittently injects system prompts with statements that reinforce an LMs initial traits and is more effective than system prompt repetition at keeping models in line with their initial pro-social behaviors. Importantly, this method does not require access to model parameters or internal model states, making it suitable for increasingly common use cases where complex multi-agent workflows are being designed with black box models.

This paper introduces the Shepherd Test, a new conceptual test for assessing the moral and relational dimensions of superintelligent artificial agents. The test is inspired by human interactions with animals, where ethical considerations about care, manipulation, and consumption arise in contexts of asymmetric power and self-preservation. We argue that AI crosses an important, and potentially dangerous, threshold of intelligence when it exhibits the ability to manipulate, nurture, and instrumentally use less intelligent agents, while also managing its own survival and expansion goals. This includes the ability to weigh moral trade-offs between self-interest and the well-being of subordinate agents. The Shepherd Test thus challenges traditional AI evaluation paradigms by emphasizing moral agency, hierarchical behavior, and complex decision-making under existential stakes. We argue that this shift is critical for advancing AI governance, particularly as AI systems become increasingly integrated into multi-agent environments. We conclude by identifying key research directions, including the development of simulation environments for testing moral behavior in AI, and the formalization of ethical manipulation within multi-agent systems.

Bandit algorithms and Large Language Models (LLMs) have emerged as powerful tools in artificial intelligence, each addressing distinct yet complementary challenges in decision-making and natural language processing. This survey explores the synergistic potential between these two fields, highlighting how bandit algorithms can enhance the performance of LLMs and how LLMs, in turn, can provide novel insights for improving bandit-based decision-making. We first examine the role of bandit algorithms in optimizing LLM fine-tuning, prompt engineering, and adaptive response generation, focusing on their ability to balance exploration and exploitation in large-scale learning tasks. Subsequently, we explore how LLMs can augment bandit algorithms through advanced contextual understanding, dynamic adaptation, and improved policy selection using natural language reasoning. By providing a comprehensive review of existing research and identifying key challenges and opportunities, this survey aims to bridge the gap between bandit algorithms and LLMs, paving the way for innovative applications and interdisciplinary research in AI.

Our paper argues that the majority of theory of mind benchmarks are broken because of their inability to directly test how large language models (LLMs) adapt to new partners. This problem stems from the fact that theory of mind benchmarks for LLMs are overwhelmingly inspired by the methods used to test theory of mind in humans and fall victim to a fallacy of attributing human-like qualities to AI agents. We expect that humans will engage in a consistent reasoning process across various questions about a situation, but this is known to not be the case for current LLMs. Most theory of mind benchmarks only measure what we call literal theory of mind: the ability to predict the behavior of others. However, this type of metric is only informative when agents exhibit self-consistent reasoning. Thus, we introduce the concept of functional theory of mind: the ability to adapt to agents in-context following a rational response to their behavior. We find that many open source LLMs are capable of displaying strong literal theory of mind capabilities, but seem to struggle with functional theory of mind -- even with exceedingly simple partner policies. Simply put, strong literal theory of mind performance does not necessarily imply strong functional theory of mind performance or vice versa. Achieving functional theory of mind, particularly over long interaction horizons with a partner, is a significant challenge deserving a prominent role in any meaningful LLM theory of mind evaluation.

Bandit algorithms and Large Language Models (LLMs) have emerged as powerful tools in artificial intelligence, each addressing distinct yet complementary challenges in decision-making and natural language processing. This survey explores the synergistic potential between these two fields, highlighting how bandit algorithms can enhance the performance of LLMs and how LLMs, in turn, can provide novel insights for improving bandit-based decision-making. We first examine the role of bandit algorithms in optimizing LLM fine-tuning, prompt engineering, and adaptive response generation, focusing on their ability to balance exploration and exploitation in large-scale learning tasks. Subsequently, we explore how LLMs can augment bandit algorithms through advanced contextual understanding, dynamic adaptation, and improved policy selection using natural language reasoning. By providing a comprehensive review of existing research and identifying key challenges and opportunities, this survey aims to bridge the gap between bandit algorithms and LLMs, paving the way for innovative applications and interdisciplinary research in AI.

The increased use of personal assistants has made question-answering a common method for user-system interaction. In these systems, while it is easy to observe implicit feedbacks such as a user clicking on a link provided by the QA system, they can be noisy. On the other hand, receiving explicit feedback on the response is rare but more valuable. To address this issue, this paper proposes a new stochastic multi-armed bandit model that considers both types of feedbacks, noisy and sparse rewards. The model is studied in both classical and contextual bandit settings, and efficient algorithm is proposed and analyzed based on the UCB framework. This algorithm is evaluated through empirical studies on various reward distributions and a real-world dataset and application.

Developing value-aligned agents is a complex undertaking and an ongoing challenge in the field of AI. Indeed, designing Large Language Models (LLMs) that can balance multiple possibly conflicting moral values based on the context is a problem of paramount importance. In this paper, we propose a system that performs contextual value alignment based on contextual aggregation of possible responses. This aggregation is achieved by integrating a subset of possible LLM responses that are best suited to a user's input while taking into account features extracted about the user's moral preferences. The proposed system trained using the Moral Integrity Corpus displays better alignment to human values than state-of-the-art baselines.

Although parameter-efficient fine-tuning methods, such as LoRA, only modify a small subset of parameters, they can have a significant impact on the model. Our instruction-tuning experiments show that LoRA-based supervised fine-tuning can catastrophically degrade model capabilities, even when trained on very small datasets for relatively few steps. With that said, we demonstrate that while the most straightforward approach (that is likely the most used in practice) fails spectacularly, small tweaks to the training procedure with very little overhead can virtually eliminate the problem. Particularly, in this paper we consider a regularized approximate replay approach which penalizes KL divergence with respect to the initial model and interleaves in data for next token prediction from a different, yet similar, open access corpus to what was used in pre-training. When applied to Qwen instruction-tuned models, we find that this recipe preserves general knowledge in the model without hindering plasticity to new tasks by adding a modest amount of computational overhead.

Much of the research focus on AI alignment seeks to align large language models and other foundation models to the context-less and generic values of helpfulness, harmlessness, and honesty. Frontier model providers also strive to align their models with these values. In this paper, we motivate why we need to move beyond such a limited conception and propose three dimensions for doing so. The first scope of alignment is competence: knowledge, skills, or behaviors the model must possess to be useful for its intended purpose. The second scope of alignment is transience: either semantic or episodic depending on the context of use. The third scope of alignment is audience: either mass, public, small-group, or dyadic. At the end of the paper, we use the proposed framework to position some technologies and workflows that go beyond prevailing notions of alignment.

The endowment of AI with reasoning capabilities and some degree of agency is widely viewed as a path toward more capable and generalizable systems. Our position is that the current development of agentic AI requires a more holistic, systems-theoretic perspective in order to fully understand their capabilities and mitigate any emergent risks. The primary motivation for our position is that AI development is currently overly focused on individual model capabilities, often ignoring broader emergent behavior, leading to a significant underestimation in the true capabilities and associated risks of agentic AI. We describe some fundamental mechanisms by which advanced capabilities can emerge from (comparably simpler) agents simply due to their interaction with the environment and other agents. Informed by an extensive amount of existing literature from various fields, we outline mechanisms for enhanced agent cognition, emergent causal reasoning ability, and metacognitive awareness. We conclude by presenting some key open challenges and guidance for the development of agentic AI. We emphasize that a systems-level perspective is essential for better understanding, and purposefully shaping, agentic AI systems.

Building pluralistic AI requires designing models that are able to be shaped to represent a wide range of value systems and cultures. Achieving this requires first being able to evaluate the degree to which a given model is capable of reflecting various personas. To this end, we propose a benchmark for evaluating the steerability of model personas as a function of prompting. Our design is based on a formal definition of prompt steerability, which analyzes the degree to which a model's joint behavioral distribution can be shifted from its baseline. By defining steerability indices and inspecting how these indices change as a function of steering effort, we can estimate the steerability of a model across various persona dimensions and directions. Our benchmark reveals that the steerability of many current models is limited - due to both a skew in their baseline behavior and an asymmetry in their steerability across many persona dimensions. We release an implementation of our benchmark at https: //github.com/IBM/prompt-steering.

The therapeutic working alliance is a critical predictor of psychotherapy success. Traditionally, working alliance assessment relies on questionnaires completed by both therapists and patients. In this paper, we present COMPASS, a novel framework to directly infer the therapeutic working alliance from the natural language used in psychotherapy sessions. Our approach leverages advanced large language models (LLMs) to analyze session transcripts and map them to distributed representations. These representations capture the semantic similarities between the dialogues and psychometric instruments, such as the Working Alliance Inventory. Analyzing a dataset of over 950 sessions spanning diverse psychiatric conditions -- including anxiety (N=498), depression (N=377), schizophrenia (N=71), and suicidal tendencies (N=12) -- collected between 1970 and 2012, we demonstrate the effectiveness of our method in providing fine-grained mapping of patient-therapist alignment trajectories, offering interpretable insights for clinical practice, and identifying emerging patterns related to the condition being treated. By employing various deep learning-based topic modeling techniques in combination with prompting generative language models, we analyze the topical characteristics of different psychiatric conditions and how these topics evolve during each turn of the conversation. This integrated framework enhances the understanding of therapeutic interactions, enables timely feedback for therapists on the quality of therapeutic relationships, and provides clear, actionable insights to improve the effectiveness of psychotherapy.