Pith. sign in

REVIEW 8 cited by

Specializing Smaller Language Models towards Multi-Step Reasoning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2301.12726 v1 pith:YA6EVEET submitted 2023-01-30 cs.CL cs.AIcs.LG

Specializing Smaller Language Models towards Multi-Step Reasoning

classification cs.CL cs.AIcs.LG
keywords modelsmodelabilityreasoningsmallertowardsabilitiesimportant
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

The surprising ability of Large Language Models (LLMs) to perform well on complex reasoning with only few-shot chain-of-thought prompts is believed to emerge only in very large-scale models (100+ billion parameters). We show that such abilities can, in fact, be distilled down from GPT-3.5 ($\ge$ 175B) to T5 variants ($\le$ 11B). We propose model specialization, to specialize the model's ability towards a target task. The hypothesis is that large models (commonly viewed as larger than 100B) have strong modeling power, but are spread on a large spectrum of tasks. Small models (commonly viewed as smaller than 10B) have limited model capacity, but if we concentrate their capacity on a specific target task, the model can achieve a decent improved performance. We use multi-step math reasoning as our testbed because it is a very typical emergent ability. We show two important aspects of model abilities: (1). there exists a very complex balance/ tradeoff between language models' multi-dimensional abilities; (2). by paying the price of decreased generic ability, we can clearly lift up the scaling curve of models smaller than 10B towards a specialized multi-step math reasoning ability. We further give comprehensive discussions about important design choices for better generalization, including the tuning data format, the start model checkpoint, and a new model selection method. We hope our practice and discoveries can serve as an important attempt towards specialized smaller models in the new research paradigm set by LLMs.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Toward Calibrated, Fair, and accurate Deepfake Detection

    cs.LG 2026-06 unverdicted novelty 7.0

    Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.

  2. LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks

    cs.LG 2026-07 conditional novelty 6.0

    Training a language model by distilling a coach's written experiential knowledge beats training on a scalar rubric score for open-ended tasks, with better out-of-distribution transfer.

  3. Scaling Relationship on Learning Mathematical Reasoning with Large Language Models

    cs.CL 2023-08 unverdicted novelty 6.0

    Pre-training loss predicts LLM math reasoning better than parameter count; rejection sampling fine-tuning with diverse paths raises LLaMA-7B accuracy on GSM8K from 35.9% with SFT to 49.3%.

  4. ReWOO: Decoupling Reasoning from Observations for Efficient Augmented Language Models

    cs.CL 2023-05 conditional novelty 6.0

    ReWOO decouples reasoning from tool observations in augmented language models, delivering 5x token efficiency and 4% higher accuracy on multi-step reasoning benchmarks like HotpotQA.

  5. Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes

    cs.CL 2023-05 conditional novelty 6.0

    Distilling step-by-step uses LLM-generated rationales as additional supervision in a multi-task framework so that 770M-parameter models outperform 540B-parameter models on NLP benchmarks with only 80% of the data.

  6. On-Policy Distillation with Best-of-N Teacher Rollout Selection

    cs.CV 2026-05 unverdicted novelty 5.0

    BRTS improves on-policy distillation by sampling multiple teacher rollouts and selecting the best one via a correctness-first then alignment priority rule, yielding gains on AIME and AMC math benchmarks.

  7. On-Policy Distillation with Best-of-N Teacher Rollout Selection

    cs.CV 2026-05 unverdicted novelty 5.0

    BRTS improves on-policy distillation by selecting the highest-quality teacher trajectory from a small pool of samples based on correctness and alignment with the student, yielding gains on AIME and AMC math benchmarks.

  8. Will LLMs Scaling Hit the Wall? Breaking Barriers via Distributed Resources on Massive Edge Devices

    cs.DC 2025-03 unverdicted novelty 2.0

    Position paper claiming that distributed training across massive edge devices can overcome data depletion and centralized compute monopolies in LLM scaling.