SCAR: Shapley Credit Assignment for More Efficient RLHF

Reinforcement Learning from Human Feedback (RLHF) is a widely used technique for aligning Large Language Models (LLMs) with human preferences, yet it often suffers from sparse reward signals, making effective credit assignment challenging. In typical setups, the reward model provides a single scalar score for an entire generated sequence, offering little insight into which token or span-level decisions were responsible for the outcome. To address this, we propose Shapley Credit Assignment Rewards (SCAR), a novel method that leverages Shapley values in cooperative game theory. SCAR distributes the total sequence-level reward among constituent tokens or text spans based on their principled marginal contributions. This creates dense reward signals, crucially, without necessitating the training of auxiliary critique models or recourse to fine-grained human annotations at intermediate generation stages. Unlike prior dense reward methods, SCAR offers a game-theoretic foundation for fair credit attribution. Theoretically, we demonstrate that SCAR preserves the original optimal policy, and empirically, across diverse tasks including sentiment control, text summarization, and instruction tuning, we show that SCAR converges significantly faster and achieves higher final reward scores compared to standard RLHF and attention-based dense reward baselines. Our findings suggest that SCAR provides a more effective and theoretically sound method for credit assignment in RLHF, leading to more efficient alignment of LLMs.

Uncovering a Universal Abstract Algorithm for Modular Addition in Neural Networks

We propose a testable universality hypothesis, asserting that seemingly disparate neural network solutions observed in the simple task of modular addition are unified under a common abstract algorithm. While prior work interpreted variations in neuron-level representations as evidence for distinct algorithms, we demonstrate - through multi-level analyses spanning neurons, neuron clusters, and entire networks - that multilayer perceptrons and transformers universally implement the abstract algorithm we call the approximate Chinese Remainder Theorem. Crucially, we introduce approximate cosets and show that neurons activate exclusively on them. Furthermore, our theory works for deep neural networks (DNNs). It predicts that universally learned solutions in DNNs with trainable embeddings or more than one hidden layer require only O(log n) features, a result we empirically confirm. This work thus provides the first theory-backed interpretation of multilayer networks solving modular addition. It advances generalizable interpretability and opens a testable universality hypothesis for group multiplication beyond modular addition.

Plasticity as the Mirror of Empowerment

Agents are minimally entities that are influenced by their past observations and act to influence future observations. This latter capacity is captured by empowerment, which has served as a vital framing concept across artificial intelligence and cognitive science. This former capacity, however, is equally foundational: In what ways, and to what extent, can an agent be influenced by what it observes? In this paper, we ground this concept in a universal agent-centric measure that we refer to as plasticity, and reveal a fundamental connection to empowerment. Following a set of desiderata on a suitable definition, we define plasticity using a new information-theoretic quantity we call the generalized directed information. We show that this new quantity strictly generalizes the directed information introduced by Massey (1990) while preserving all of its desirable properties. Our first finding is that plasticity is the mirror of empowerment: The agent's plasticity is identical to the empowerment of the environment, and vice versa. Our second finding establishes a tension between the plasticity and empowerment of an agent, suggesting that agent design needs to be mindful of both characteristics. We explore the implications of these findings, and suggest that plasticity, empowerment, and their relationship are essential to understanding agency.

Language Agents Mirror Human Causal Reasoning Biases. How Can We Help Them Think Like Scientists?

Language model (LM) agents are increasingly used as autonomous decision-makers who need to actively gather information to guide their decisions. A crucial cognitive skill for such agents is the efficient exploration and understanding of the causal structure of the world -- key to robust, scientifically grounded reasoning. Yet, it remains unclear whether LMs possess this capability or exhibit systematic biases leading to erroneous conclusions. In this work, we examine LMs' ability to explore and infer causal relationships, using the well-established "Blicket Test" paradigm from developmental psychology. We find that LMs reliably infer the common, intuitive disjunctive causal relationships but systematically struggle with the unusual, yet equally (or sometimes even more) evidenced conjunctive ones. This "disjunctive bias" persists across model families, sizes, and prompting strategies, and performance further declines as task complexity increases. Interestingly, an analogous bias appears in human adults, suggesting that LMs may have inherited deep-seated reasoning heuristics from their training data. To this end, we quantify similarities between LMs and humans, finding that LMs exhibit adult-like inference profiles (but not children-like). Finally, we propose a test-time sampling method which explicitly samples and eliminates hypotheses about causal relationships from the LM. This scalable approach significantly reduces the disjunctive bias and moves LMs closer to the goal of scientific, causally rigorous reasoning.

Cracking the Code of Action: a Generative Approach to Affordances for Reinforcement Learning

Agents that can autonomously navigate the web through a graphical user interface (GUI) using a unified action space (e.g., mouse and keyboard actions) can require very large amounts of domain-specific expert demonstrations to achieve good performance. Low sample efficiency is often exacerbated in sparse-reward and large-action-space environments, such as a web GUI, where only a few actions are relevant in any given situation. In this work, we consider the low-data regime, with limited or no access to expert behavior. To enable sample-efficient learning, we explore the effect of constraining the action space through $\textit{intent-based affordances}$ -- i.e., considering in any situation only the subset of actions that achieve a desired outcome. We propose $\textbf{Code as Generative Affordances}$ $(\textbf{$\texttt{CoGA}$})$, a method that leverages pre-trained vision-language models (VLMs) to generate code that determines affordable actions through implicit intent-completion functions and using a fully-automated program generation and verification pipeline. These programs are then used in-the-loop of a reinforcement learning agent to return a set of affordances given a pixel observation. By greatly reducing the number of actions that an agent must consider, we demonstrate on a wide range of tasks in the MiniWob++ benchmark that: $\textbf{1)}$ $\texttt{CoGA}$ is orders of magnitude more sample efficient than its RL agent, $\textbf{2)}$ $\texttt{CoGA}$'s programs can generalize within a family of tasks, and $\textbf{3)}$ $\texttt{CoGA}$ performs better or on par compared with behavior cloning when a small number of expert demonstrations is available.

Capturing Individual Human Preferences with Reward Features

Reinforcement learning from human feedback usually models preferences using a reward model that does not distinguish between people. We argue that this is unlikely to be a good design choice in contexts with high potential for disagreement, like in the training of large language models. We propose a method to specialise a reward model to a person or group of people. Our approach builds on the observation that individual preferences can be captured as a linear combination of a set of general reward features. We show how to learn such features and subsequently use them to quickly adapt the reward model to a specific individual, even if their preferences are not reflected in the training data. We present experiments with large language models comparing the proposed architecture with a non-adaptive reward model and also adaptive counterparts, including models that do in-context personalisation. Depending on how much disagreement there is in the training data, our model either significantly outperforms the baselines or matches their performance with a simpler architecture and more stable training.

Agency Is Frame-Dependent

Agency is a system's capacity to steer outcomes toward a goal, and is a central topic of study across biology, philosophy, cognitive science, and artificial intelligence. Determining if a system exhibits agency is a notoriously difficult question: Dennett (1989), for instance, highlights the puzzle of determining which principles can decide whether a rock, a thermostat, or a robot each possess agency. We here address this puzzle from the viewpoint of reinforcement learning by arguing that agency is fundamentally frame-dependent: Any measurement of a system's agency must be made relative to a reference frame. We support this claim by presenting a philosophical argument that each of the essential properties of agency proposed by Barandiaran et al. (2009) and Moreno (2018) are themselves frame-dependent. We conclude that any basic science of agency requires frame-dependence, and discuss the implications of this claim for reinforcement learning.

Langevin Soft Actor-Critic: Efficient Exploration through Uncertainty-Driven Critic Learning

Existing actor-critic algorithms, which are popular for continuous control reinforcement learning (RL) tasks, suffer from poor sample efficiency due to lack of principled exploration mechanism within them. Motivated by the success of Thompson sampling for efficient exploration in RL, we propose a novel model-free RL algorithm, Langevin Soft Actor Critic (LSAC), which prioritizes enhancing critic learning through uncertainty estimation over policy optimization. LSAC employs three key innovations: approximate Thompson sampling through distributional Langevin Monte Carlo (LMC) based $Q$ updates, parallel tempering for exploring multiple modes of the posterior of the $Q$ function, and diffusion synthesized state-action samples regularized with $Q$ action gradients. Our extensive experiments demonstrate that LSAC outperforms or matches the performance of mainstream model-free RL algorithms for continuous control tasks. Notably, LSAC marks the first successful application of an LMC based Thompson sampling in continuous control tasks with continuous action spaces.

Fairness in Reinforcement Learning with Bisimulation Metrics

Ensuring long-term fairness is crucial when developing automated decision making systems, specifically in dynamic and sequential environments. By maximizing their reward without consideration of fairness, AI agents can introduce disparities in their treatment of groups or individuals. In this paper, we establish the connection between bisimulation metrics and group fairness in reinforcement learning. We propose a novel approach that leverages bisimulation metrics to learn reward functions and observation dynamics, ensuring that learners treat groups fairly while reflecting the original problem. We demonstrate the effectiveness of our method in addressing disparities in sequential decision making problems through empirical evaluation on a standard fairness benchmark consisting of lending and college admission scenarios.

MaestroMotif: Skill Design from Artificial Intelligence Feedback

Describing skills in natural language has the potential to provide an accessible way to inject human knowledge about decision-making into an AI system. We present MaestroMotif, a method for AI-assisted skill design, which yields high-performing and adaptable agents. MaestroMotif leverages the capabilities of Large Language Models (LLMs) to effectively create and reuse skills. It first uses an LLM's feedback to automatically design rewards corresponding to each skill, starting from their natural language description. Then, it employs an LLM's code generation abilities, together with reinforcement learning, for training the skills and combining them to implement complex behaviors specified in language. We evaluate MaestroMotif using a suite of complex tasks in the NetHack Learning Environment (NLE), demonstrating that it surpasses existing approaches in both performance and usability.