Salvato in:
Dettagli Bibliografici
Autori principali: Fang, Jiaxin, He, Runyuan, Bhatia, Sahil, Gajare, Neel, Cheung, Alvin
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2604.16931
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914486413164544
author Fang, Jiaxin
He, Runyuan
Bhatia, Sahil
Gajare, Neel
Cheung, Alvin
author_facet Fang, Jiaxin
He, Runyuan
Bhatia, Sahil
Gajare, Neel
Cheung, Alvin
contents Recent advances in large language models (LLMs) have shown that test-time scaling can substantially improve model performance on complex tasks, particularly in the coding domain. Under this paradigm, models use a larger token budget during inference to generate intermediate reasoning traces before producing a final answer. However, current evaluations primarily rely on competitive programming benchmarks, which may not capture the full range of reasoning abilities. In this work, we perform a systematic study of frontier reasoning models to understand their performance on real-world coding benchmarks. To gain more insights into the performance of such models, we devise a programmatic way to {\em automatically generate} coding tasks of arbitrary difficulty and structure from existing benchmarks. Using this framework, our analysis reveals that the structure of a reasoning trace, not just its contents, is a strong predictor of correctness. Motivated by this, we propose structured thought-trees as means to represent reasoning traces. To illustrate their use, we train a lightweight classifier on features extracted from thought-trees to predict trace correctness, and demonstrate that flagging and retrying structurally anomalous traces based on the extracted features yields consistent gains at lower complexity levels.
format Preprint
id arxiv_https___arxiv_org_abs_2604_16931
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Playing Psychic: Using Thought Trees to Predict Reasoning Models Accuracy on Coding Tasks
Fang, Jiaxin
He, Runyuan
Bhatia, Sahil
Gajare, Neel
Cheung, Alvin
Artificial Intelligence
Recent advances in large language models (LLMs) have shown that test-time scaling can substantially improve model performance on complex tasks, particularly in the coding domain. Under this paradigm, models use a larger token budget during inference to generate intermediate reasoning traces before producing a final answer. However, current evaluations primarily rely on competitive programming benchmarks, which may not capture the full range of reasoning abilities. In this work, we perform a systematic study of frontier reasoning models to understand their performance on real-world coding benchmarks. To gain more insights into the performance of such models, we devise a programmatic way to {\em automatically generate} coding tasks of arbitrary difficulty and structure from existing benchmarks. Using this framework, our analysis reveals that the structure of a reasoning trace, not just its contents, is a strong predictor of correctness. Motivated by this, we propose structured thought-trees as means to represent reasoning traces. To illustrate their use, we train a lightweight classifier on features extracted from thought-trees to predict trace correctness, and demonstrate that flagging and retrying structurally anomalous traces based on the extracted features yields consistent gains at lower complexity levels.
title Playing Psychic: Using Thought Trees to Predict Reasoning Models Accuracy on Coding Tasks
topic Artificial Intelligence
url https://arxiv.org/abs/2604.16931