# Textual Self-attention Network: Test-Time Preference Optimization through Textual Gradient-based Attention

arXiv: https://arxiv.org/abs/2511.06682 · HTML: https://mlsanity.com/paper/2511.06682/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 23 of its 36 references that are in the map.
- First posted on arXiv 10 Nov 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [LLM preference optimization methods](https://mlsanity.com/alignment/llm-preference-optimization-methods/)

## Builds on (most cited first, 23 of 23)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluators](https://mlsanity.com/paper/2404.04475/) (2024)
- [BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset](https://mlsanity.com/paper/2307.04657/) (2023)
- [ORPO: Monolithic Preference Optimization without Reference Model](https://mlsanity.com/paper/2403.07691/) (2024)
- [Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation](https://mlsanity.com/paper/2401.08417/) (2024)
- [RLHF Workflow: From Reward Modeling to Online RLHF](https://mlsanity.com/paper/2405.07863/) (2024)
- [Chain of Preference Optimization: Improving Chain-of-Thought Reasoning in LLMs](https://mlsanity.com/paper/2406.09136/) (2024)
- [Reinforcement Learning for LLM Post-Training: A Survey](https://mlsanity.com/paper/2407.16216/) (2024)
- [Aligning LLMs with Individual Preferences via Interaction](https://mlsanity.com/paper/2410.03642/) (2024)
- [RMB: Comprehensively Benchmarking Reward Models in LLM Alignment](https://mlsanity.com/paper/2410.09893/) (2024)
- [TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree-Search and Best-of-N Sampling](https://mlsanity.com/paper/2410.16033/) (2024)
- [Test-Time Preference Optimization: On-the-Fly Alignment via Iterative Textual Feedback](https://mlsanity.com/paper/2501.12895/) (2025)
- [Data-Efficient Alignment of Large Language Models with Human Feedback Through Natural Language](https://mlsanity.com/paper/2311.14543/) (2023)
- [A Survey on LLM Inference-Time Self-Improvement](https://mlsanity.com/paper/2412.14352/) (2024)
- [Rethinking the Role of Proxy Rewards in Language Model Alignment](https://mlsanity.com/paper/2402.03469/) (2024)
- [Training Language Model to Critique for Better Refinement](https://mlsanity.com/paper/2506.22157/) (2025)

## Abstract (arXiv)

Large Language Models (LLMs) have demonstrated remarkable generalization capabilities, but aligning their outputs with human preferences typically requires expensive supervised fine-tuning. Recent test-time methods leverage textual feedback to overcome this, but they often critique and revise a single candidate response, lacking a principled mechanism to systematically analyze, weigh, and synthesize the strengths of multiple promising candidates. Such a mechanism is crucial because different responses may excel in distinct aspects (e.g., clarity, factual accuracy, or tone), and combining their best elements may produce a far superior outcome. This paper proposes the Textual Self-Attention Network (TSAN), a new paradigm for test-time preference optimization that requires no parameter updates. TSAN emulates self-attention entirely in natural language to overcome this gap: it analyzes multiple candidates by formatting them into textual keys and values, weighs their relevance using an LLM-based attention module, and synthesizes their strengths into a new, preference-aligned response under the guidance of the learned textual attention. This entire process operates in a textual gradient space, enabling iterative and interpretable optimization. Empirical evaluations demonstrate that with just three test-time iterations on a base SFT model, TSAN outperforms supervised models like Llama-3.1-70B-Instruct and surpasses the current state-of-the-art test-time alignment method by effectively leveraging multiple candidate solutions.