AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Probing How Scalable Table Data Enhances General Long-Context Reasoning

Huaibing Xie, Guoliang Zhao, Yang Liu, Shihan Dou, Siming Huang, Yanling Xiao, Shaolei Wang, Yiting Liu, Cheng Zhang, Shaofan Liu, Pluto Zhou

ICML 2026 · Large language models › Language, safety and interpretability · approach: Controlled text generation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
RL on synthesized table data (TableLong)DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 40.12 -> 48.36Improved
RL on TableLong dataDS-R1-Distill-14B · 7 long-context benchmarks averageaverage accuracy: 36.14 -> 45.07Improved
RL on TableLong dataDS-R1-Distill-32B · Needle in a Haystackretrieval accuracy (%): 87.95 -> 99.40Improved
RL on TableLong dataDS-R1-Distill-32B · LiveCodeBench, AIME 2025, GPQA-Diamond, MultiChallenge (OOD)accuracy: LCB 46.71 -> 58.68; AIME 60.00 -> 70.00Improved
RL on TableLong dataQwen3-32B · MRCRaccuracy: 42.45 -> 42.66No clear effect
RL on TableLong dataDS-R1-Distill-14B · Loong / GSM-Infinite / MRCRaccuracy: Loong 25.07 -> 35.44; GSM-INF 9.00 -> 9.60Mixed
use semantics-free (structure only) tables (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 40.12 -> 41.79 (vs 48.36 with full data)Mixed
add cell semantics to table data (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 41.79 -> 48.36Improved
add visible delimiters to tables (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 47.59 -> 48.36No clear effect
replace delimiters with noise text (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 48.36 -> 47.38No clear effect
increase number of involved table cells (0~30 to 300+) (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 46.30 -> 48.36Improved
increase number of tables per instance (1 to 1~30) (ablation)TableLong, DS-R1-Distill-32B · 7 long-context benchmarks averageaverage accuracy: 46.66 -> 48.36Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy