# DatedGPT: Preventing Lookahead Bias in Large Language Models with Time-Aware Pretraining

arXiv: https://arxiv.org/abs/2603.11838 · HTML: https://mlsanity.com/paper/2603.11838/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 39 references that are in the map.
- First posted on arXiv 12 Mar 2026.

## Builds on (most cited first, 22 of 22)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [Enriching Word Vectors with Subword Information](https://mlsanity.com/paper/1607.04606/) (2016)
- [Get To The Point: Summarization with Pointer-Generator Networks](https://mlsanity.com/paper/1704.04368/) (2017)
- [The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale](https://mlsanity.com/paper/2406.17557/) (2024)
- [WizardLM: Empowering large pre-trained language models to follow complex instructions](https://mlsanity.com/paper/2304.12244/) (2023)
- [TinyLlama: An Open-Source Small Language Model](https://mlsanity.com/paper/2401.02385/) (2024)
- [OLMo: Accelerating the Science of Language Models](https://mlsanity.com/paper/2402.00838/) (2024)
- [SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents](https://mlsanity.com/paper/1611.04230/) (2016)
- [MAmmoTH2: Scaling Instructions from the Web](https://mlsanity.com/paper/2405.03548/) (2024)
- [Stop Uploading Test Data in Plain Text: Practical Strategies for Mitigating Data Contamination by Evaluation Benchmarks](https://mlsanity.com/paper/2305.10160/) (2023)
- [The Art of Saying No: Contextual Noncompliance in Language Models](https://mlsanity.com/paper/2407.12043/) (2024)
- [Meta-Learning Online Adaptation of Language Models](https://mlsanity.com/paper/2305.15076/) (2023)
- [Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection](https://mlsanity.com/paper/2502.06042/) (2025)
- [Assessing Look-Ahead Bias in Stock Return Predictions Generated By GPT Sentiment Analysis](https://mlsanity.com/paper/2309.17322/) (2023)
- [Infini-gram mini: Exact n-gram Search at the Internet Scale with FM-Index](https://mlsanity.com/paper/2506.12229/) (2025)

## Built on by (most cited first, 8 of 8)

- [Scaling Point-in-Time Language Models: Enabling Economic Evaluation of Embeddings](https://mlsanity.com/paper/2607.11889/) (2026)
- [A Language Model from 1913: Pretraining on Historical Text](https://mlsanity.com/paper/2606.02991/) (2026)
- [Temporal Leakage in LLM Backtesting: Measurement, Validation, and Adjusted Scores](https://mlsanity.com/paper/2608.02985/) (2026)
- [Hindcast: Replaying Prediction Markets to Evaluate LLM Forecasters](https://mlsanity.com/paper/2607.14051/) (2026)
- [The Self Driving Portfolio: Agentic Architecture for Institutional Asset Management](https://mlsanity.com/paper/2604.02279/) (2026)
- [HindsightBench: A Black-Box Behavioral Audit Protocol for Parametric Hindsight in Time-Indexed LLM Decision Tasks](https://mlsanity.com/paper/2607.18867/) (2026)
- [SciTBERT: A family of chronologically consistent language models for scientific and technological language processing](https://mlsanity.com/paper/2610.12207/) (2026)
- [L"ushi-1840: Building and Probing a Knowledge Cutoff for a Classical Chinese Language Model](https://mlsanity.com/paper/2610.16201/) (2026)

## Abstract (arXiv)

Large language models pretrained on internet-scale data risk lookahead bias in forecasting tasks, as they may have already seen the true outcome during training. To address this, we present DatedGPT, a family of twelve 1.3B-parameter language models trained from scratch on approximately 100 billion tokens each with strict annual data cutoffs spanning 2013 to 2024, together with DatedInstruct, an instruction dataset grounded in each year's documents to prevent leakage during post-training. The models are competitive with open models of similar scale, and perplexity-based probing confirms that each model's knowledge is bounded by its cutoff year. On stock return prediction over 61,000 firm-day news headlines, DatedGPT-instruct achieves an annualised Sharpe ratio of 3.20 under the lookahead-bias-free setup. Lookahead-biased models, whose training data covers the outcome period, add a lookahead premium of 26.4 b.p. per standard deviation, significant at the 1% level. The series thus enables direct analysis of lookahead bias in financial forecasting. We provide an interactive web demo that allows users to query and compare responses from models across different cutoff years, available at www.datedgpt.com.