Revealing Behavioral Plasticity in Large Language Models: A Token-Conditional Perspective

• URL: http://arxiv.org/abs/2603.08398v1
• Date: Mon, 09 Mar 2026 13:56:53 GMT
• Title: Revealing Behavioral Plasticity in Large Language Models: A Token-Conditional Perspective
• Abstract summary: We show that Large Language Models (LLMs) possess intrinsic behavioral plasticity that can be exposed through token-conditional generation and stabilized via reinforcement learning.<n>We propose Token-Conditioned Reinforcement Learning (ToCoRL) to transform transient inference-time adaptations into stable and learnable behavioral patterns.
• Score: 63.9266662672413
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In this work, we reveal that Large Language Models (LLMs) possess intrinsic behavioral plasticity-akin to chameleons adapting their coloration to environmental cues-that can be exposed through token-conditional generation and stabilized via reinforcement learning. Specifically, by conditioning generation on carefully selected token prefixes sampled from responses exhibiting desired behaviors, LLMs seamlessly adapt their behavioral modes at inference time (e.g., switching from step-by-step reasoning to direct answering) without retraining. Based on this insight, we propose Token-Conditioned Reinforcement Learning (ToCoRL), a principled framework that leverages RL to internalize this chameleon-like plasticity, transforming transient inference-time adaptations into stable and learnable behavioral patterns. ToCoRL guides exploration with token-conditional generation and keep enhancing exploitation, enabling emergence of appropriate behaviors. Extensive experiments show that ToCoRL enables precise behavioral control without capability degradation. Notably, we show that large reasoning models, while performing strongly on complex mathematics, can be effectively adapted to excel at factual question answering, which was a capability previously hindered by their step-by-step reasoning patterns.

Related papers

• On Distinguishing Capability Elicitation from Capability Creation in Post-Training: A Free-Energy Perspective [13.996919933596153]
  We argue that post-training research should distinguish between capability elicitation and capability creation.<n>We develop this argument through a free-energy view of post-training.
  arXiv  Detail & Related papers  (2026-05-08T18:23:25Z)
• LLMs Reading the Rhythms of Daily Life: Aligned Understanding for Behavior Prediction and Generation [53.62804271492357]
  Large language models (LLMs) offer a promising direction due to their semantic richness, strong interpretability, and generative capabilities.<n>We propose Behavior Understanding Alignment (BUA), a novel framework that integrates LLMs into human behavior modeling through a structured curriculum learning process.<n>BUA employs sequence embeddings from pretrained behavior models as alignment anchors and guides the LLM through a three-stage curriculum, while a multi-round dialogue setting introduces prediction and generation capabilities.
  arXiv  Detail & Related papers  (2026-04-26T07:34:37Z)
• PromptCD: Test-Time Behavior Enhancement via Polarity-Prompt Contrastive Decoding [85.22047087898311]
  We introduce Polarity-Prompt Contrastive Decoding (PromptCD), a test-time behavior control method that generalizes contrastive decoding to broader enhancement settings.<n>PromptCD constructs paired positive and negative guiding prompts for a target behavior and contrasts model responses to reinforce desirable outcomes.<n>Experiments on the "3H" alignment objectives demonstrate consistent and substantial improvements, indicating that post-trained models can achieve meaningful self-enhancement purely at test time.
  arXiv  Detail & Related papers  (2026-02-24T08:56:52Z)
• Reflection Pretraining Enables Token-Level Self-Correction in Biological Sequence Models [82.79223371188756]
  Chain-of-Thought (CoT) prompting has advanced task-solving capabilities in natural language processing with large language models.<n>Applying CoT to non-natural language domains, such as protein and RNA language models, is not yet possible.<n>We introduce pretraining, for the first time in a biological sequence model, which enables the model to engage in intermediate reasoning.
  arXiv  Detail & Related papers  (2025-12-24T05:25:17Z)
• Reasoning Palette: Modulating Reasoning via Latent Contextualization for Controllable Exploration for (V)LMs [49.66344956133349]
  Reasoning capacity shapes both inference-time performance and reinforcement learning (RL) training for large (vision-) language models.<n>This paper proposes Reasoning Palette, a novel latent-modulation framework that endows the model with a latent variable for strategic contextualization.
  arXiv  Detail & Related papers  (2025-12-19T03:32:53Z)
• Can Large Reasoning Models Self-Train? [51.0277533541394]
  We use majority voting as a simple self-feedback mechanism to study whether self-training can be sustained within reinforcement learning.<n>We find that this basic approach improves not only the model's reasoning performance, but also its capability of generating better quality feedback for the next RL iteration.<n>Yet our analysis also reveals a critical limitation of such a self-training paradigm - prolonged RL with self-reward leads to reward hacking, resulting in sudden and complete performance collapse.
  arXiv  Detail & Related papers  (2025-05-27T17:16:00Z)
• Extended Inductive Reasoning for Personalized Preference Inference from Behavioral Signals [45.019257216564036]
  This paper investigates extended inductive reasoning in large language models (LLMs)<n>We propose AlignXplore, a model that enables systematic preference inference from behavioral signals in users' interaction histories.<n>We show that AlignXplore achieves substantial improvements over the backbone model by an average of 15.49% on in-domain and out-of-domain benchmarks.
  arXiv  Detail & Related papers  (2025-05-23T16:16:46Z)
• Predictability Shapes Adaptation: An Evolutionary Perspective on Modes of Learning in Transformers [51.992454203752686]
  Transformer models learn in two distinct modes: in-weights learning (IWL) and in-context learning (ICL)<n>We draw inspiration from evolutionary biology's analogous adaptive strategies: genetic encoding and phenotypic plasticity.<n>We experimentally operationalize these dimensions of predictability and investigate their influence on the ICL/IWL balance in Transformers.
  arXiv  Detail & Related papers  (2025-05-14T23:31:17Z)
• Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs [28.565225092457897]
  Reinforcement learning can drive self-improvement in language models on verifiable tasks.<n>We find that Qwen-2.5-3B far exceeds Llama-3.2-3B under identical RL training for the game of Countdown.<n>Our study reveals that Qwen naturally exhibits these reasoning behaviors, whereas Llama initially lacks them.
  arXiv  Detail & Related papers  (2025-03-03T08:46:22Z)
• REACT: Revealing Evolutionary Action Consequence Trajectories for Interpretable Reinforcement Learning [7.889696505137217]
  We propose Revealing Evolutionary Action Consequence Trajectories (REACT) to enhance the interpretability of Reinforcement Learning (RL) In contrast to the prevalent practice of RL models based on their optimal behavior learned during training, we posit that considering a range of edge-case trajectories provides a more comprehensive understanding of their inherent behavior. Our results highlight its effectiveness in revealing nuanced aspects of RL models' behavior beyond optimal performance, thereby contributing to improved interpretability.
  arXiv  Detail & Related papers  (2024-04-04T10:56:30Z)
• Learning Complex Spatial Behaviours in ABM: An Experimental Observational Study [0.0]
  This paper explores how Reinforcement Learning can be applied to create emergent agent behaviours. Running a series of simulations, we demonstrate that agents trained using the novel Proximal Policy optimisation algorithm behave in ways that exhibit properties of real-world intelligent adaptive behaviours.
  arXiv  Detail & Related papers  (2022-01-04T11:56:11Z)

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.