# CEDAR-GRPO: Process-Aware Reinforcement Learning for General Abductive Reasoning in LLMs

arXiv: https://arxiv.org/abs/2608.14791 · HTML: https://mlsanity.com/paper/2608.14791/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 22 of its 43 references that are in the map.
- First posted on arXiv 14 Aug 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 22 of 22)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams](https://mlsanity.com/paper/2009.13081/) (2020)
- [Solving math word problems with process- and outcome-based feedback](https://mlsanity.com/paper/2211.14275/) (2022)
- [Abductive Commonsense Reasoning](https://mlsanity.com/paper/1908.05739/) (2019)
- [MuSR: Testing the Limits of Chain-of-thought with Multistep Soft Reasoning](https://mlsanity.com/paper/2310.16049/) (2023)
- [CLIMATE-FEVER: A Dataset for Verification of Real-World Climate Claims](https://mlsanity.com/paper/2012.00614/) (2020)
- [DDXPlus: A New Dataset For Automatic Medical Diagnosis](https://mlsanity.com/paper/2205.09148/) (2022)
- [Are Large Language Models Really Good Logical Reasoners? A Comprehensive Evaluation and Beyond](https://mlsanity.com/paper/2306.09841/) (2023)
- [ArgMed-Agents: Explainable Clinical Decision Reasoning with LLM Disscusion via Argumentation Schemes](https://mlsanity.com/paper/2403.06294/) (2024)
- [Patterns Over Principles: The Fragility of Inductive Reasoning in LLMs under Noisy Observations](https://mlsanity.com/paper/2502.16169/) (2025)
- [Controllable Logical Hypothesis Generation for Abductive Reasoning in Knowledge Graphs](https://mlsanity.com/paper/2505.20948/) (2025)
- [IFAR: Multi-Perspective and Multi-Level Causal Discovery with LLMs](https://mlsanity.com/paper/2409.05559/) (2024)
- [Abductive Inference in Retrieval-Augmented Language Models: Generating and Validating Missing Premises](https://mlsanity.com/paper/2511.04020/) (2025)
- [Unifying Deductive and Abductive Reasoning in Knowledge Graphs with Masked Diffusion Model](https://mlsanity.com/paper/2510.11462/) (2025)
- [An Incomplete Loop: Deductive, Inductive, and Abductive Reasoning in Language Models](https://mlsanity.com/paper/2404.03028/) (2024)
- [GEAR: A General Evaluation Framework for Abductive Reasoning](https://mlsanity.com/paper/2509.24096/) (2025)
- [Wiring the 'Why': A Unified Taxonomy and Survey of Abductive Reasoning in LLMs](https://mlsanity.com/paper/2604.08016/) (2026)
- [From Hypothesis to Premises: LLM-based Backward Logical Reasoning with Selective Symbolic Translation](https://mlsanity.com/paper/2512.03360/) (2025)

## Abstract (arXiv)

Abductive reasoning, often characterized as inference to the best explanation, is central to explanation under uncertainty, from everyday sense-making and investigation to scientific discovery. Yet LLM research has mostly studied abduction through narrow, task-specific benchmarks, making it unclear whether observed gains transfer beyond the benchmark family used for training or evaluation. We ask whether RL post-training can improve abduction as a transferable reasoning capability. We introduce CEDAR-GRPO, a process-aware framework that combines final-answer correctness with abductive rewards for evidence coverage and evidence-to-explanation directionality. Four open-weight LLMs are post-trained on a controlled, domain-neutral mixture of abductive hypothesis-generation and hypothesis-selection tasks. We evaluate them on 11 unseen tasks spanning hypothesis selection, missing-fact generation, defeasible inference, long-context investigation, clinical reasoning, code debugging, and non-abductive controls. CEDAR- GRPO improves every model on every held-out task over both base models and correctness-only GRPO, with average gains of 7.4 and 2.7 points, respectively, and a maximum gain of 30.8 points. Ablations confirm that RL, abductive reward design, and task diversity each contribute to transfer. Process-level metrics further show stronger abductive behavior, including exploration of alternatives, elimination of rivals, backtracking, and uncertainty marking.