---
title: "ARCHER: From Rationale Imitation to... | SI Paper Desk"
url: https://thesimind.si/papers/2611.18494/
description: "ARCHER trains small SI models to solve math by writing step-by-step traces that call a calculator and check results, and the authors report gains over the..."
---

Number 4 of 12 · November 17, 2026

# ARCHER: From Rationale Imitation to Executable Reasoning in Small Language Models

ARCHER trains small SI models to solve math by writing step-by-step traces that call a calculator and check results, and the authors report gains over the strongest tool-augmented baseline.

Dhruv Dixit, Khushi Patel, Congcong Xu and 1 more · Submitted November 13, 2026 · arXiv 2611.18494v1 · cs.AI

[Read the paper on arXiv](https://arxiv.org/abs/2611.18494) · [The PDF](https://arxiv.org/pdf/2611.18494) · [The day's 12](https://thesimind.si/papers/day/2026-11-17/)

## What they found

The authors built ARCHER, a two-stage way to train small language models (under 10B parameters) for math. In stage one, free-form chain-of-thought explanations are turned into typed action traces. These traces hold reasoning steps, calculator calls, results from a deterministic runtime, verification actions, and a final answer. In stage two, reinforcement learning refines the model using rewards for correct answers, tool grounding, process discipline, and trace alignment. Across five standard sub-10B backbones, ARCHER beats the strongest tool-augmented baseline by 5.51 percentage points on macro-average accuracy. Gains include 8.95 points on MATH-500, and 5.33 and 5.34 points on AIME-2024 and AIME-2025. The authors also report benefits on two more 7B backbones that were already trained for reasoning.

## Why it matters

Small models are cheap to run, can work locally, and can keep data private. But they often fail at exact math. This work suggests that teaching them to use a calculator and check their own work, instead of imitating explanations, can make them more reliable. The traces are also easier to audit, since each step and tool result is recorded.

## How it changes the race

If the results hold, smaller teams could get stronger math reasoning from compact models without needing very large ones. It looks like a solid step for tool-grounded training of small models, not a change in frontier capability.

## Safety and control

The auditable traces, with recorded tool calls and verification steps, could make a model's reasoning easier for people to inspect. The abstract does not test safety or misuse directly.

### The abstract, by the authors

Small language models are attractive for local, low-cost, and privacy-sensitive deployment, but they remain unreliable on mathematical reasoning tasks that require decomposition, exact computation, tool use, and verification. We introduce ARCHER: Action-trace Reinforcement for Calculator-grounded Hierarchical Executable Reasoning, a two-stage framework for training small language models (SLMs) to produce executable and auditable reasoning trajectories. In the first stage, ARCHER compiles free-form chain-of-thought rationales into typed action traces containing reasoning steps, calculator calls, deterministic runtime observations, verification actions, and final answers. This shifts the supervision target from imitating explanatory text to learning a tool-grounded reasoning policy. In the second stage, ARCHER refines this policy through asymmetric hierarchical reinforcement learning, using grouped and conditionally gated trajectory rewards for answer correctness, tool grounding, process discipline, and trace alignment. Across five standard sub-10B backbones, ARCHER improves over the strongest tool-augmented baseline by 5.51 percentage points on macro-average across benchmark accuracies, including gains of 8.95 points on MATH-500 and 5.33 and 5.34 points on AIME-2024 and AIME-2025. Evaluations on two additional reasoning-specialized 7B backbones further show that executable grounding remains beneficial after reasoning-focused reinforcement learning or distillation. These results demonstrate that executable supervision and reward-guided behavioral refinement provide a general and effective training paradigm for compact mathematical reasoners.

Quoted from arXiv:2611.18494. The research and its words belong to its authors.

Desk score **77** out of 100

Why it was picked

- Reasoning (in the title)
- Alignment
- Agents
- Evals
- Language models (in the title)

Topics

- [Reasoning](https://thesimind.si/papers/?tag=reasoning)
- [Math](https://thesimind.si/papers/?tag=math)
- [Reinforcement learning](https://thesimind.si/papers/?tag=reinforcement-learning)
- [Agents](https://thesimind.si/papers/?tag=agents)
- [Efficiency](https://thesimind.si/papers/?tag=efficiency)
- [Language models](https://thesimind.si/papers/?tag=language-models)

**arXiv ID:** 2611.18494v1
**Submitted:** November 13, 2026
**Fields:** cs.AI (Artificial Intelligence)
**Level:** Intermediate

Fingerprint, SHA-256

`f6717e7aa1609d6cf8e88326d9eefff87222dbbf6fb4334a8ba6b707151cb402`

On Solana: [Proof of Mind seal, 2HNJWs...cNY9](https://solscan.io/tx/2HNJWsLa5Tp3F7wQaNr1CPR71N1aoGNptA7fgCafC5m3BwbmKwhNGTXyaSfyEa9jqug6kM55u5VmDvbSrRepcNY9)

### The exact text fingerprinted

{"arxiv":"2611.18494","version":1,"title":"ARCHER: From Rationale Imitation to Executable Reasoning in Small Language Models","authors":["Dhruv Dixit","Khushi Patel","Congcong Xu","Min Song"],"day":"2026-11-17","rank":4,"score":77,"status":"written","summary":{"takeaway":"ARCHER trains small SI models to solve math by writing step-by-step traces that call a calculator and check results, and the authors report gains over the strongest tool-augmented baseline.","found":"The authors built ARCHER, a two-stage way to train small language models (under 10B parameters) for math. In stage one, free-form chain-of-thought explanations are turned into typed action traces. These traces hold reasoning steps, calculator calls, results from a deterministic runtime, verification actions, and a final answer. In stage two, reinforcement learning refines the model using rewards for correct answers, tool grounding, process discipline, and trace alignment. Across five standard sub-10B backbones, ARCHER beats the strongest tool-augmented baseline by 5.51 percentage points on macro-average accuracy. Gains include 8.95 points on MATH-500, and 5.33 and 5.34 points on AIME-2024 and AIME-2025. The authors also report benefits on two more 7B backbones that were already trained for reasoning.","matters":"Small models are cheap to run, can work locally, and can keep data private. But they often fail at exact math. This work suggests that teaching them to use a calculator and check their own work, instead of imitating explanations, can make them more reliable. The traces are also easier to audit, since each step and tool result is recorded.","race":"If the results hold, smaller teams could get stronger math reasoning from compact models without needing very large ones. It looks like a solid step for tool-grounded training of small models, not a change in frontier capability.","safety":"The auditable traces, with recorded tool calls and verification steps, could make a model's reasoning easier for people to inspect. The abstract does not test safety or misuse directly.","level":"intermediate","tags":["reasoning","math","reinforcement-learning","agents","efficiency","language-models"]}}

[Check it on Verify a thought](https://thesimind.si/proof-of-mind/)

## The authors

Dhruv Dixit, Khushi Patel, Congcong Xu, Min Song

Credit for this research belongs to its authors. SI MIND's page is a plain-English reading of their abstract on arXiv, not a review of the full paper.

Related on the desk

## More papers on these topics

[Nov 18, 2026 · Number 7GPRM: Global Perspective Process Reward ModelGPRM is a step-by-step reasoning judge that sees earlier steps, their scores, and later steps, and the authors report it beats strong baselines on PRMBench and ProcessBench with less data and smaller models.ReasoningEvalsReinforcement learning](https://thesimind.si/papers/2611.20755/) · [Nov 20, 2026 · Number 9QDA: Closed-Loop LLM Quantization Design with AgentsQDA uses an SI agent to design the compression recipe for each language model, and the abstract says it beats the strongest public recipes on four reasoning models.EfficiencyAgentsLanguage models](https://thesimind.si/papers/2611.23388/) · [Nov 15, 2026 · Number 6Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon AgentsA SI model that is bad at acting in a long task can still be a good teacher, because judging actions after the episode ends shows their real consequences.AgentsReinforcement learningTraining](https://thesimind.si/papers/2611.17167/)

[Number 3](https://thesimind.si/papers/2611.17821/) · [All 12 for Nov 17, 2026](https://thesimind.si/papers/day/2026-11-17/) · [Number 5](https://thesimind.si/papers/2611.18514/)
