# Cross-Environment Neural Reranking for Sample-Efficient Action Selection in Text-Based Agents

arXiv: https://arxiv.org/abs/2606.02204 · HTML: https://mlsanity.com/paper/2606.02204/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use.
- Builds on 15 of its 18 references that are in the map.
- First posted on arXiv 1 Jun 2026.

## Placed in

- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 15 of 15)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Gradient Surgery for Multi-Task Learning](https://mlsanity.com/paper/2001.06782/) (2020)
- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents](https://mlsanity.com/paper/2207.01206/) (2022)
- [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020)
- [Recurrent World Models Facilitate Policy Evolution](https://mlsanity.com/paper/1809.01999/) (2018)
- [ScienceWorld: Is your Agent Smarter than a 5th Grader?](https://mlsanity.com/paper/2203.07540/) (2022)
- [Web Agents with World Models: Learning and Leveraging Environment Dynamics in Web Navigation](https://mlsanity.com/paper/2410.13232/) (2024)
- [Agent Planning with World Knowledge Model](https://mlsanity.com/paper/2405.14205/) (2024)
- [JudgeRank: Leveraging Large Language Models for Reasoning-Intensive Reranking](https://mlsanity.com/paper/2411.00142/) (2024)
- [Why Do LLM Agents Fail in Exploring New Environments? A World-Modeling Perspective](https://mlsanity.com/paper/2510.15047/) (2025)
- [CoPS: Empowering LLM Agents with Provable Cross-Task Experience Sharing](https://mlsanity.com/paper/2410.16670/) (2024)
- [Retrospex: Language Agent Meets Offline Reinforcement Learning Critic](https://mlsanity.com/paper/2505.11807/) (2025)
- [EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning](https://mlsanity.com/paper/2403.10692/) (2024)

## Built on by (most cited first, 1 of 1)

- [What Can Component-Replacement Evidence Establish? A Critical Scoping Review of Local Decisions in LLM Agents](https://mlsanity.com/paper/2609.39989/) (2026)

## Abstract (arXiv)

Large language model agents achieve strong performance on text-based benchmarks but incur prohibitive inference costs, motivating the use of compact neural rerankers for action selection. We investigate whether a single lightweight model can perform action selection across multiple diverse environments, a capability that would eliminate per-environment model maintenance. Training DeBERTa-v3 (184M-434M parameters) jointly on ALFWorld, WebShop, and ScienceWorld with minority-class upsampling, we find that rebalanced two-environment joint training substantially improves over single-environment ALFWorld performance (net gain +0.412) while maintaining competitive WebShop performance (+0.214 vs. +0.249 single-environment). Three-environment training yields a mean combined net gain of +0.551 +/- 0.024 across 4 seeds, with per-environment results approaching specialized single-environment models while providing positive cross-domain transfer. Cross-environment adaptation is highly sample-efficient: fine-tuning on only 9.2% of target-domain data recovers 93% of full-data performance, and scaling model capacity yields limited benefits, indicating data diversity is the primary driver. Environment-aware LoRA adapter routing with PCGrad achieves a best-seed result of +0.611 (seed 42), with seeds 456 and 789 at +0.554 and +0.559, but exhibits high variance due to seed 123 collapsing to +0.263 (4-seed mean +0.497 +/- 0.158), representing a promising but currently unstable direction. Joint training with clean splits and data rebalancing is a key ingredient. We will release our three-environment benchmark of 51,580 training instances (41,740 raw unique states with minority-class upsampling) and all model checkpoints upon acceptance.