Pith. sign in

REVIEW 11 cited by

Scaling Laws of RoPE-based Extrapolation

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 2310.05209 v2 pith:64D6T63R submitted 2023-10-08 cs.CL cs.AI

Scaling Laws of RoPE-based Extrapolation

classification cs.CL cs.AI
keywords extrapolationlengthrope-basedbasecontextfine-tuninglargerlaws
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

The extrapolation capability of Large Language Models (LLMs) based on Rotary Position Embedding is currently a topic of considerable interest. The mainstream approach to addressing extrapolation with LLMs involves modifying RoPE by replacing 10000, the rotary base of $\theta_n={10000}^{-2n/d}$ in the original RoPE, with a larger value and providing longer fine-tuning text. In this work, we first observe that fine-tuning a RoPE-based LLM with either a smaller or larger base in pre-training context length could significantly enhance its extrapolation performance. After that, we propose \textbf{\textit{Scaling Laws of RoPE-based Extrapolation}}, a unified framework from the periodic perspective, to describe the relationship between the extrapolation performance and base value as well as tuning context length. In this process, we also explain the origin of the RoPE-based extrapolation issue by \textbf{\textit{critical dimension for extrapolation}}. Besides these observations and analyses, we achieve extrapolation up to 1 million context length within only 16K training length on LLaMA2 7B and 13B.

discussion (0)

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

Forward citations

Cited by 11 Pith papers

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

  1. RoPE Distinguishes Neither Positions Nor Tokens in Long Contexts, Provably

    cs.CL 2026-05 conditional novelty 7.0

    Proves that RoPE attention loses locality bias and token distinction in long contexts, approaching random behavior independent of content.

  2. LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens

    cs.CL 2024-02 unverdicted novelty 7.0

    LongRoPE extends LLM context windows to 2048k tokens via search for non-uniform positional interpolation, progressive fine-tuning from 256k, and short-context readjustment.

  3. Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE

    cs.LG 2026-07 accept novelty 6.0

    Dynamic bifocal RoPE with adaptive group size recovers the base model at short context and outperforms fixed zero-shot long-context methods up to 128K with near-FA2 cost.

  4. PJ-RoPE: A Fourier-Jet-Affine Position Space for Relative Attention

    cs.LG 2026-06 unverdicted novelty 6.0

    PJ-RoPE organizes relative-position mechanisms as a learnable Fourier-Jet-Affine space derived from lag-shift dynamics, extending RoPE and ALiBi with explicit jets and sector selection.

  5. Remember to Forget: Gated Adaptive Positional Encoding

    cs.LG 2026-05 unverdicted novelty 6.0

    GAPE augments RoPE with query- and key-dependent gates to stabilize attention and improve long-context performance in language models.

  6. FocuSFT: Bilevel Optimization for Dilution-Aware Long-Context Fine-Tuning

    cs.CL 2026-05 unverdicted novelty 6.0

    FocuSFT uses an inner optimization loop to adapt fast-weight parameters into a parametric memory that sharpens attention on relevant content, then conditions outer-loop supervised fine-tuning on this representation, y...

  7. HyMem: Hybrid Memory Architecture with Dynamic Retrieval Scheduling

    cs.AI 2026-02 unverdicted novelty 6.0

    HyMem introduces dual-granular memory storage with a lightweight summary module for fast responses and selective activation of a deep LLM module for complex queries, outperforming full-context baselines by 92.6% lower...

  8. Positional Encoding via Token-Aware Phase Attention

    cs.CL 2025-09 unverdicted novelty 6.0

    TAPA adds a learnable phase function to attention to preserve long-range token interactions, enabling direct continual pretraining, length extrapolation, lower perplexity, and stronger retrieval than RoPE-style methods.

  9. MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent

    cs.CL 2025-07 unverdicted novelty 6.0

    MemAgent uses multi-conversation RL to train a memory agent that reads text in segments and overwrites memory, extrapolating from 8K training to 3.5M token QA with under 5% loss and 95%+ on 512K RULER.

  10. Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE

    cs.LG 2026-07 accept novelty 5.0

    Jet-Long is a tuning-free bifocal RoPE method that dynamically sets remote group size from sequence length, recovering the base model within the pretrained window and beating prior zero-shot extenders on RULER, HELMET...

  11. InternLM2 Technical Report

    cs.CL 2024-03 unverdicted novelty 5.0

    InternLM2 is a new open-source LLM that outperforms prior versions on 30 benchmarks and long-context tasks through scaled pre-training to 32k tokens and a conditional online RLHF alignment strategy.