AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

MrRoPE: Mixed-radix Rotary Position Embedding

Qingyuan Tian, Wenhong Zhu, Xiaoran Liu, Xiaofeng Wang, Rui Wang

ICLR 2026 · Large language models › Instruction and fine-tuning · approach: Fine-tuning for reasoning data · 2 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
use MrRoPE-Pro instead of YaRNLLaMA3-8B-Instruct · Proofpile PPL 128Kperplexity: 2.38 -> 2.34Improved
use MrRoPE-Pro instead of MrRoPE-UniLLaMA3-8B-Instruct · Proofpile PPL 128Kperplexity: 2.41 -> 2.34Improved
use MrRoPE-Uni instead of YaRN at the longest contextLLaMA3-8B-Instruct · Proofpile PPL 128Kperplexity: 2.38 -> 2.41Hurt
use MrRoPE-Pro instead of YaRN at short contextLLaMA2-7B-chat · Proofpile PPL 4Kperplexity: 6.02 -> 5.72Improved
use MrRoPE-Pro instead of YaRNLLaMA3-8B-Instruct · RULER 128Kscore: 79.9 -> 86.6Improved
use MrRoPE-Pro instead of YaRNQwen2.5-3B-Instruct · RULER 128Kscore: 50.1 -> 53.2Improved
use MrRoPE-Pro instead of YaRNLLaMA3-8B · Infinite-Bench KV Retrievalaccuracy: 9% -> 27%Improved
use MrRoPE-Pro instead of YaRNLLaMA3-8B · Infinite-Bench QA Dialogueaccuracy: 10% -> 22%Improved
use MrRoPE-Pro instead of YaRNLLaMA3-8B · Infinite-Bench averageaccuracy: 44% -> 49.8%Improved
use MrRoPE-Pro instead of YaRNLLaMA3-8B · Infinite-Bench Code Debugaccuracy: 3% -> 3%No clear effect
use MrRoPE-Pro instead of YaRNLLaMA2-7B-Instruct · LongBench-v2 LSDaccuracy: 30.0 -> 40.0Improved
use MrRoPE-Pro instead of YaRNLLaMA2-7B-Instruct · LongBench-v2 SDaccuracy: 16.7 -> 16.7No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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