# Learning What to Skip: Counterfactual Credit Assignment for Efficient Multi-Agent LLM Workflows

arXiv: https://arxiv.org/abs/2609.30734 · HTML: https://mlsanity.com/paper/2609.30734/

## Key facts

- No citing papers in the map yet (data through 7 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 14 of its 16 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 14 of 14)

- [Improving Factuality and Reasoning in Language Models through Multiagent Debate](https://mlsanity.com/paper/2305.14325/) (2023)
- [Selective Classification for Deep Neural Networks](https://mlsanity.com/paper/1705.08500/) (2017)
- [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024)
- [Mixture-of-Agents Enhances Large Language Model Capabilities](https://mlsanity.com/paper/2406.04692/) (2024)
- [SelectiveNet: A Deep Neural Network with an Integrated Reject Option](https://mlsanity.com/paper/1901.09192/) (2019)
- [DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference](https://mlsanity.com/paper/2004.12993/) (2020)
- [AutoMix: Automatically Mixing Language Models](https://mlsanity.com/paper/2310.12963/) (2023)
- [MasRouter: Learning to Route LLMs for Multi-Agent Systems](https://mlsanity.com/paper/2502.11133/) (2025)
- [Rethinking the Value of Multi-Agent Workflow: A Strong Single Agent Baseline](https://mlsanity.com/paper/2601.12307/) (2026)
- [GraphPlanner: Graph Memory-Augmented Agentic Routing for Multi-Agent LLMs](https://mlsanity.com/paper/2604.23626/) (2026)
- [Stochastic Self-Organization in Multi-Agent Systems](https://mlsanity.com/paper/2510.00685/) (2025)
- [Exact Is Easier: Credit Assignment for Cooperative LLM Agents](https://mlsanity.com/paper/2603.06859/) (2026)
- [Scoring Verifiers: Evaluating Synthetic Verification for Code and Reasoning](https://mlsanity.com/paper/2502.13820/) (2025)
- [CARD: Towards Conditional Design of Multi-agent Topological Structures](https://mlsanity.com/paper/2603.01089/) (2026)

## Abstract (arXiv)

Multi-agent LLM workflows use planning, execution, verification, and summarization to improve task performance, yet the value of each component depends on the state already produced. Executing every component can waste computation or overwrite a correct intermediate answer. We formulate component omission as counterfactual credit assignment: full-workflow logs reveal the executed trajectory's reward, while controlled skip interventions reveal the consequences of omitting a future step. We introduce Learning What to Skip (LW2S), which learns action-specific safety models from these interventions and combines held-out calibration with domain-native guards to select skips. When an early skip is rejected, the controller can continue execution and reconsider a later component. Across mathematical reasoning, multiple-choice QA, and code generation with two instruction-model families, LW2S reduces recorded token cost while matching or improving aggregate full-workflow accuracy in the evaluated settings. Scale-up and second-topology experiments further examine component redundancy, while shared-error cases reveal why agreement alone is insufficient for skip selection. These findings connect efficient workflow execution to learning the conditional utility of individual components.