Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information

• URL: http://arxiv.org/abs/2502.14258v1
• Date: Thu, 20 Feb 2025 04:52:05 GMT
• Title: Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information
• Authors: Yein Park, Chanwoong Yoon, Jungwoo Park, Minbyul Jeong, Jaewoo Kang,
• Abstract summary: Temporal Heads are specific attention heads primarily responsible for processing temporal knowledge through circuit analysis. We confirm that these heads are present across multiple models, though their specific locations may vary. We expand the potential of our findings by demonstrating how temporal knowledge can be edited by adjusting the values of these heads.
• Score: 16.28488243884373
• License:
• Abstract: While the ability of language models to elicit facts has been widely investigated, how they handle temporally changing facts remains underexplored. We discover Temporal Heads, specific attention heads primarily responsible for processing temporal knowledge through circuit analysis. We confirm that these heads are present across multiple models, though their specific locations may vary, and their responses differ depending on the type of knowledge and its corresponding years. Disabling these heads degrades the model's ability to recall time-specific knowledge while maintaining its general capabilities without compromising time-invariant and question-answering performances. Moreover, the heads are activated not only numeric conditions ("In 2004") but also textual aliases ("In the year ..."), indicating that they encode a temporal dimension beyond simple numerical representation. Furthermore, we expand the potential of our findings by demonstrating how temporal knowledge can be edited by adjusting the values of these heads.

Related papers

• XForecast: Evaluating Natural Language Explanations for Time Series Forecasting [72.57427992446698]
  Time series forecasting aids decision-making, especially for stakeholders who rely on accurate predictions. Traditional explainable AI (XAI) methods, which underline feature or temporal importance, often require expert knowledge. evaluating forecast NLEs is difficult due to the complex causal relationships in time series data.
  arXiv  Detail & Related papers  (2024-10-18T05:16:39Z)
• ChroKnowledge: Unveiling Chronological Knowledge of Language Models in Multiple Domains [19.428141279030527]
  Large language models (LLMs) have brought significant changes to many aspects of our lives. Existing approaches fall short in addressing the temporal adaptability of knowledge. We present ChroKnowledge, a novel sampling-based framework for evaluating LLMs' non-parametric chronological knowledge.
  arXiv  Detail & Related papers  (2024-10-13T15:08:49Z)
• Time Awareness in Large Language Models: Benchmarking Fact Recall Across Time [0.0]
  In real-world scenarios, the correctness of answers is frequently tied to temporal context. We present a novel framework and dataset spanning over 8,000 events from 2018 to 2024. Our work provides a significant step toward advancing time-aware language models.
  arXiv  Detail & Related papers  (2024-09-20T08:57:20Z)
• Retrieval Head Mechanistically Explains Long-Context Factuality [56.78951509492645]
  We show that a special type of attention heads are largely responsible for retrieving information, which we dub retrieval heads. We show that retrieval heads strongly influence chain-of-thought (CoT) reasoning, where the model needs to frequently refer back the question and previously-generated context. We believe our insights will foster future research on reducing hallucination, improving reasoning, and compressing the KV cache.
  arXiv  Detail & Related papers  (2024-04-24T00:24:03Z)
• Remember This Event That Year? Assessing Temporal Information and Reasoning in Large Language Models [1.472789264981363]
  Large Language Models (LLMs) are increasingly ubiquitous, yet their ability to retain and reason about temporal information remains limited. Our study experiments with 12 state-of-the-art models on a novel numerical-temporal dataset, textbfTempUN, spanning from 10,000 BCE to 2100 CE.
  arXiv  Detail & Related papers  (2024-02-19T09:43:03Z)
• Mitigating Temporal Misalignment by Discarding Outdated Facts [58.620269228776294]
  Large language models are often used under temporal misalignment, tasked with answering questions about the present. We propose fact duration prediction: the task of predicting how long a given fact will remain true. Our data and code are released publicly at https://github.com/mikejqzhang/mitigating_misalignment.
  arXiv  Detail & Related papers  (2023-05-24T07:30:08Z)
• RECKONING: Reasoning through Dynamic Knowledge Encoding [51.076603338764706]
  We show that language models can answer questions by reasoning over knowledge provided as part of the context. In these situations, the model fails to distinguish the knowledge that is necessary to answer the question. We propose teaching the model to reason more robustly by folding the provided contextual knowledge into the model's parameters.
  arXiv  Detail & Related papers  (2023-05-10T17:54:51Z)
• The KITMUS Test: Evaluating Knowledge Integration from Multiple Sources in Natural Language Understanding Systems [87.3207729953778]
  We evaluate state-of-the-art coreference resolution models on our dataset. Several models struggle to reason on-the-fly over knowledge observed both at pretrain time and at inference time. Still, even the best performing models seem to have difficulties with reliably integrating knowledge presented only at inference time.
  arXiv  Detail & Related papers  (2022-12-15T23:26:54Z)
• A Dataset for Answering Time-Sensitive Questions [88.95075983560331]
  Time is an important dimension in our physical world. Lots of facts can evolve with respect to time. It is important to consider the time dimension and empower the existing QA models to reason over time. The existing QA datasets contain rather few time-sensitive questions, hence not suitable for diagnosing or benchmarking the model's temporal reasoning capability.
  arXiv  Detail & Related papers  (2021-08-13T16:42:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.