# Triviality Corrected Endogenous Reward

arXiv: https://arxiv.org/abs/2604.11522 · HTML: https://mlsanity.com/paper/2604.11522/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 35 references that are in the map.
- First posted on arXiv 13 Apr 2026.

## Builds on (most cited first, 25 of 25)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [Solving Quantitative Reasoning Problems with Language Models](https://mlsanity.com/paper/2206.14858/) (2022)
- [Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?](https://mlsanity.com/paper/2504.13837/) (2025)
- [Can Large Language Models Be an Alternative to Human Evaluations?](https://mlsanity.com/paper/2305.01937/) (2023)
- [Learning to Reason under Off-Policy Guidance](https://mlsanity.com/paper/2504.14945/) (2025)
- [Learning to Reason without External Rewards](https://mlsanity.com/paper/2505.19590/) (2025)
- [LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs](https://mlsanity.com/paper/2408.07055/) (2024)
- [Right Question is Already Half the Answer: Fully Unsupervised LLM Reasoning Incentivization](https://mlsanity.com/paper/2504.05812/) (2025)
- [On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification](https://mlsanity.com/paper/2508.05629/) (2025)
- [An Empirical Study of LLM-as-a-Judge for LLM Evaluation: Fine-tuned Judge Model is not a General Substitute for GPT-4](https://mlsanity.com/paper/2403.02839/) (2024)
- [Jointly Reinforcing Diversity and Quality in Language Model Generations](https://mlsanity.com/paper/2509.02534/) (2025)
- [NoveltyBench: Evaluating Language Models for Humanlike Diversity](https://mlsanity.com/paper/2504.05228/) (2025)
- [Confidence Is All You Need: Few-Shot RL Fine-Tuning of Language Models](https://mlsanity.com/paper/2506.06395/) (2025)
- [Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling](https://mlsanity.com/paper/2507.01679/) (2025)
- [Base Models Beat Aligned Models at Randomness and Creativity](https://mlsanity.com/paper/2505.00047/) (2025)
- [Reverse-Engineered Reasoning for Open-Ended Generation](https://mlsanity.com/paper/2509.06160/) (2025)
- [Are We on the Right Way to Assessing LLM-as-a-Judge?](https://mlsanity.com/paper/2512.16041/) (2025)
- [RL Is Neither a Panacea Nor a Mirage: Understanding Supervised vs. Reinforcement Learning Fine-Tuning for LLMs](https://mlsanity.com/paper/2508.16546/) (2025)
- [RLAC: Reinforcement Learning with Adversarial Critic for Free-Form Generation Tasks](https://mlsanity.com/paper/2511.01758/) (2025)
- [Igniting Creative Writing in Small Language Models: LLM-as-a-Judge versus Multi-Agent Refined Rewards](https://mlsanity.com/paper/2508.21476/) (2025)
- [Can Reasoning Help Large Language Models Capture Human Annotator Disagreement?](https://mlsanity.com/paper/2506.19467/) (2025)
- [Writing as a testbed for open ended agents](https://mlsanity.com/paper/2503.19711/) (2025)
- [Align to Structure: Aligning Large Language Models with Structural Information](https://mlsanity.com/paper/2504.03622/) (2025)
- [EvolvR: Self-Evolving Pairwise Reasoning for Story Evaluation to Enhance Generation](https://mlsanity.com/paper/2508.06046/) (2025)

## Abstract (arXiv)

Reinforcement learning for open-ended text generation is constrained by the lack of verifiable rewards, necessitating reliance on judge models that require either annotated data or powerful closed-source models. Inspired by recent work on unsupervised reinforcement learning for mathematical reasoning using confidence-based endogenous rewards, we investigate whether this principle can be adapted to open-ended writing tasks. We find that directly applying confidence rewards leads to Triviality Bias: the policy collapses toward high-probability outputs, reducing diversity and meaningful content. We propose TCER (Triviality Corrected Endogenous Reward), which addresses this bias by rewarding the relative information gain between a specialist policy and a generalist reference policy, modulated by a probability-dependent correction mechanism. Across multiple writing benchmarks and model architectures, TCER achieves consistent improvements without external supervision. Furthermore, TCER also transfers effectively to mathematical reasoning, validating the generality of our approach across different generation tasks.