# OCC-RAG: Optimal Cognitive Core for Faithful Question Answering

arXiv: https://arxiv.org/abs/2606.00683 · HTML: https://mlsanity.com/paper/2606.00683/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning.
- Builds on 26 of its 34 references that are in the map.
- First posted on arXiv 30 May 2026.

## Placed in

- LLM reasoning / [Knowledge- and tool-augmented reasoning](https://mlsanity.com/lr/knowledge-and-tool-augmented-reasoning/)

## Builds on (most cited first, 26 of 26)

- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [MuSiQue: Multihop Questions via Single-hop Question Composition](https://mlsanity.com/paper/2108.00573/) (2021)
- [It's Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners](https://mlsanity.com/paper/2009.07118/) (2020)
- [TAT-QA: A Question Answering Benchmark on a Hybrid of Tabular and Textual Content in Finance](https://mlsanity.com/paper/2105.07624/) (2021)
- [Small Language Models are the Future of Agentic AI](https://mlsanity.com/paper/2506.02153/) (2025)
- [AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions](https://mlsanity.com/paper/2506.09038/) (2025)
- [xLAM: A Family of Large Action Models to Empower AI Agent Systems](https://mlsanity.com/paper/2409.03215/) (2024)
- [FaithEval: Can Your Language Model Stay Faithful to Context, Even If "The Moon is Made of Marshmallows"](https://mlsanity.com/paper/2410.03727/) (2024)
- [TinyGSM: achieving >80% on GSM8k with small language models](https://mlsanity.com/paper/2312.09241/) (2023)
- [Evaluating Step-by-step Reasoning Traces: A Survey](https://mlsanity.com/paper/2502.12289/) (2025)
- [Natural Language Reasoning, A Survey](https://mlsanity.com/paper/2303.14725/) (2023)
- [Context-DPO: Aligning Language Models for Context-Faithfulness](https://mlsanity.com/paper/2412.15280/) (2024)
- [Characterizing LLM Abstention Behavior in Science QA with Context Perturbations](https://mlsanity.com/paper/2404.12452/) (2024)
- [Task Matters: Knowledge Requirements Shape LLM Responses to Context-Memory Conflict](https://mlsanity.com/paper/2506.06485/) (2025)
- [Task-Specific Efficiency Analysis: When Small Language Models Outperform Large Language Models](https://mlsanity.com/paper/2603.21389/) (2026)
- [When Models Lie, We Learn: Multilingual Span-Level Hallucination Detection with PsiloQA](https://mlsanity.com/paper/2510.04849/) (2025)
- [Quantifying reliance on external information over parametric knowledge during Retrieval Augmented Generation (RAG) using mechanistic analysis](https://mlsanity.com/paper/2410.00857/) (2024)
- [When Silence Is Golden: Can LLMs Learn to Abstain in Temporal QA and Beyond?](https://mlsanity.com/paper/2602.04755/) (2026)
- [Wikontic: Constructing Wikidata-Aligned, Ontology-Aware Knowledge Graphs with Large Language Models](https://mlsanity.com/paper/2512.00590/) (2025)
- [Will It Still Be True Tomorrow? Multilingual Evergreen Question Classification to Improve Trustworthy QA](https://mlsanity.com/paper/2505.21115/) (2025)
- [DRAGOn: Designing RAG On Periodically Updated Corpus](https://mlsanity.com/paper/2507.05713/) (2025)
- [Even Small Reasoners Should Quote Their Sources: Introducing the Pleias-RAG Model Family](https://mlsanity.com/paper/2504.18225/) (2025)
- [SIN-Bench: Tracing Native Evidence Chains in Long-Context Multimodal Scientific Interleaved Literature](https://mlsanity.com/paper/2601.10108/) (2026)
- [Long-Document QA with Chain-of-Structured-Thought and Fine-Tuned SLMs](https://mlsanity.com/paper/2603.29232/) (2026)
- [On Synthesizing Data for Context Attribution in Question Answering](https://mlsanity.com/paper/2504.05317/) (2025)

## Abstract (arXiv)

Recent progress in the development of language models has been defined by scale, with each generation absorbing more of the world's knowledge into its weights. However, many practical applications benefit more from robust reasoning than from extensive parametric knowledge. In this setting, task-specialized small language models (SLMs) offer a principled design choice. We introduce Optimal Cognitive Core (OCC), a family of SLMs built around this premise. As a variant of OCC, we present OCC-RAG, optimized for faithful question answering (QA) grounded in the provided context. This task directly aligns with the OCC design approach, requiring multi-hop reasoning over supplied passages while ignoring memorized knowledge. To train OCC-RAG, we implement a novel pipeline for synthesizing multi-context, multi-hop QA data at scale, producing a corpus of over three million examples targeting multi-hop reasoning, strict context faithfulness, and calibrated abstention. We release OCC-RAG-0.6B and OCC-RAG-1.7B, both mid-trained on this corpus. The models produce structured reasoning traces with source citations grounded in literal quotes from the context. Through OCC-RAG, we demonstrate that compact, task-specialized SLMs can match or exceed general-purpose models 2 -- 6x their size across multi-hop reasoning (HotpotQA, MuSiQue, TAT-QA), faithfulness (ConFiQA), and refusal (MuSiQue-Un) benchmarks.