# Forgetting Curve: A Reliable Method for Evaluating Memorization Capability for Long-context Models

arXiv: https://arxiv.org/abs/2410.04727 · HTML: https://mlsanity.com/paper/2410.04727/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 40 references that are in the map.
- First posted on arXiv 7 Oct 2024.

## Builds on (most cited first, 27 of 27)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://mlsanity.com/paper/1901.02860/) (2019)
- [LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding](https://mlsanity.com/paper/2308.14508/) (2023)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [RULER: What's the Real Context Size of Your Long-Context Language Models?](https://mlsanity.com/paper/2404.06654/) (2024)
- [RWKV: Reinventing RNNs for the Transformer Era](https://mlsanity.com/paper/2305.13048/) (2023)
- [Extending Context Window of Large Language Models via Positional Interpolation](https://mlsanity.com/paper/2306.15595/) (2023)
- [Compressive Transformers for Long-Range Sequence Modelling](https://mlsanity.com/paper/1911.05507/) (2019)
- [MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies](https://mlsanity.com/paper/2404.06395/) (2024)
- [Long Range Arena: A Benchmark for Efficient Transformers](https://mlsanity.com/paper/2011.04006/) (2020)
- [ELI5: Long Form Question Answering](https://mlsanity.com/paper/1907.09190/) (2019)
- [Jamba: A Hybrid Transformer-Mamba Language Model](https://mlsanity.com/paper/2403.19887/) (2024)
- [∞Bench: Extending Long Context Evaluation Beyond 100K Tokens](https://mlsanity.com/paper/2402.13718/) (2024)
- [Long-context LLMs Struggle with Long In-context Learning](https://mlsanity.com/paper/2404.02060/) (2024)
- [L-Eval: Instituting Standardized Evaluation for Long Context Language Models](https://mlsanity.com/paper/2307.11088/) (2023)
- [Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs](https://mlsanity.com/paper/2402.03927/) (2024)
- [LooGLE: Can Long-Context Language Models Understand Long Contexts?](https://mlsanity.com/paper/2311.04939/) (2023)
- [Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention](https://mlsanity.com/paper/2404.07143/) (2024)
- [ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understanding](https://mlsanity.com/paper/2305.14196/) (2023)
- [Scaling Transformer to 1M tokens and beyond with RMT](https://mlsanity.com/paper/2304.11062/) (2023)
- [LV-Eval: A Balanced Long-Context Benchmark with 5 Length Levels Up to 256K](https://mlsanity.com/paper/2402.05136/) (2024)
- [Can Perplexity Reflect Large Language Model's Ability in Long Text Understanding?](https://mlsanity.com/paper/2405.06105/) (2024)
- [ConDefects: A New Dataset to Address the Data Leakage Concern for LLM-based Fault Localization and Program Repair](https://mlsanity.com/paper/2310.16253/) (2023)

## Built on by (most cited first, 13 of 13)

- [Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors](https://mlsanity.com/paper/2510.08907/) (2025)
- [COMI: Coarse-to-fine Context Compression via Marginal Information Gain](https://mlsanity.com/paper/2602.01719/) (2026)
- [ParamMute: Suppressing Knowledge-Critical FFNs for Faithful Retrieval-Augmented Generation](https://mlsanity.com/paper/2502.15543/) (2025)
- [ExpandR: Teaching Dense Retrievers Beyond Queries with LLM Guidance](https://mlsanity.com/paper/2502.17057/) (2025)
- [Read As Human: Compressing Context via Parallelizable Close Reading and Skimming](https://mlsanity.com/paper/2602.01840/) (2026)
- [Shared Path: Unraveling Memorization in Multilingual LLMs through Language Similarities](https://mlsanity.com/paper/2505.15722/) (2025)
- [Chunks as Arms: Multi-Armed Bandit-Guided Sampling for Long-Context LLM Preference Optimization](https://mlsanity.com/paper/2508.13993/) (2025)
- [Can an LLM Induce a Graph? Investigating Memory Drift and Context Length](https://mlsanity.com/paper/2510.03611/) (2025)
- [PoC: Performance-oriented Context Compression for Large Language Models via Performance Prediction](https://mlsanity.com/paper/2603.19733/) (2026)
- [Beyond Position Bias: Shifting Context Compression from Position-Driven to Semantic-Driven](https://mlsanity.com/paper/2605.09463/) (2026)
- [Data Distribution Matters: A Data-Centric Perspective on Context Compression for Large Language Model](https://mlsanity.com/paper/2602.01778/) (2026)
- [ForgetBench: Benchmarking Forgetting Dynamics of Long-Term Parametric Memory in Language Models](https://mlsanity.com/paper/2607.26455/) (2026)
- [Industry-Aligned Granular Topic Modeling](https://mlsanity.com/paper/2601.11762/) (2026)

## Abstract (arXiv)

Numerous recent works target to extend effective context length for language models and various methods, tasks and benchmarks exist to measure model's effective memorization length. However, through thorough investigations, we find limitations for currently existing evaluations on model's memorization capability. We provide an extensive survey for limitations in this work and propose a new method called forgetting curve to measure the memorization capability of long-context models. We show that forgetting curve has the advantage of being robust to the tested corpus and the experimental settings, of not relying on prompts and can be applied to any model size. We apply our forgetting curve to a large variety of models involving both transformer and RNN/SSM based architectures. Our measurement provides empirical evidence for the effectiveness of transformer extension techniques while raises questions for the effective length of RNN/SSM based models. We also examine the difference between our measurement and existing benchmarks as well as popular metrics for various models. Our code and results can be found at https://github.com/1azybug/ForgettingCurve.