English
Related papers

Related papers: Geometric Factual Recall in Transformers

200 papers

How do transformer-based large language models (LLMs) store and retrieve knowledge? We focus on the most basic form of this task -- factual recall, where the model is tasked with explicitly surfacing stored facts in prompts of form `Fact:…

Machine Learning · Computer Science 2024-02-14 Bilal Chughtai , Alan Cooney , Neel Nanda

In recent years, pre-trained Transformers have dominated the majority of NLP benchmark tasks. Many variants of pre-trained Transformers have kept breaking out, and most focus on designing different pre-training objectives or variants of…

Computation and Language · Computer Science 2020-10-13 Yu-An Wang , Yun-Nung Chen

Transformers excel through content-addressable retrieval and the ability to exploit contexts of, in principle, unbounded length. We recast associative memory at the level of probability measures, treating a context as a distribution over…

Machine Learning · Statistics 2026-02-03 Ryotaro Kawata , Taiji Suzuki

Factual incorrectness in generated content is one of the primary concerns in ubiquitous deployment of large language models (LLMs). Prior findings suggest LLMs can (sometimes) detect factual incorrectness in their generated content (i.e.,…

Computation and Language · Computer Science 2025-05-28 Hovhannes Tamoyan , Subhabrata Dutta , Iryna Gurevych

Large language models demonstrate remarkable ability in factual recall, yet the fundamental limits of storing and retrieving input--output associations with neural networks remain unclear. We study these limits in a minimal setting: a…

Machine Learning · Statistics 2026-05-12 Alessio Giorlandino , Sebastian Goldt , Antoine Maillard

Transformers often appear to perform Bayesian reasoning in context, but verifying this rigorously has been impossible: natural data lack analytic posteriors, and large models conflate reasoning with memorization. We address this by…

Machine Learning · Computer Science 2026-05-19 Naman Agarwal , Siddhartha R. Dalal , Vishal Misra

Pretrained embeddings based on the Transformer architecture have taken the NLP community by storm. We show that they can mathematically be reframed as a sum of vector factors and showcase how to use this reframing to study the impact of…

Computation and Language · Computer Science 2022-06-09 Timothee Mickus , Denis Paperno , Mathieu Constant

The ability of machine learning models to store input information in hidden layer vector embeddings, analogous to the concept of `memory', is widely employed but not well characterized. We find that language model embeddings typically…

Computation and Language · Computer Science 2026-05-20 Benjamin L. Badger

Language models draw on two knowledge sources: facts baked into weights (parametric memory, PM) and information in context (working memory, WM). We study two mechanistically distinct failure modes--conflict, when PM and WM disagree and…

Artificial Intelligence · Computer Science 2026-05-15 Qiyao Liang , Risto Miikkulainen , Ila Fiete

As deep generative models have progressed, recent work has shown them to be capable of memorizing and reproducing training datapoints when deployed. These findings call into question the usability of generative models, especially in light…

A few models have tried to tackle the link prediction problem, also known as knowledge graph completion, by embedding knowledge graphs in comparably lower dimensions. However, the state-of-the-art results are attained at the cost of…

Machine Learning · Computer Science 2022-11-29 Peyman Baghershahi , Reshad Hosseini , Hadi Moradi

Many models learn representations of knowledge graph data by exploiting its low-rank latent structure, encoding known relations between entities and enabling unknown facts to be inferred. To predict whether a relation holds between…

Machine Learning · Computer Science 2021-01-19 Carl Allen , Ivana Balažević , Timothy Hospedales

Representing words by vectors, or embeddings, enables computational reasoning and is foundational to automating natural language tasks. For example, if word embeddings of similar words contain similar values, word similarity can be readily…

Computation and Language · Computer Science 2022-02-02 Carl Allen

Modern approaches to enhancing Large Language Models' factual accuracy and knowledge utilization face a fundamental trade-off: non-parametric retrieval-augmented generation (RAG) provides flexible access to external knowledge but suffers…

Computation and Language · Computer Science 2026-03-02 Rubin Wei , Jiaqi Cao , Jiarui Wang , Jushi Kai , Qipeng Guo , Bowen Zhou , Zhouhan Lin

The storage and recall of factual associations in auto-regressive transformer language models (LMs) have drawn a great deal of attention, inspiring knowledge editing by directly modifying the located model weights. Most editing works…

Computation and Language · Computer Science 2025-02-28 Xiyu Liu , Zhengxiao Liu , Naibin Gu , Zheng Lin , Wanli Ma , Ji Xiang , Weiping Wang

Embeddings are a basic initial feature extraction step in many machine learning models, particularly in natural language processing. An embedding attempts to map data tokens to a low-dimensional space where similar tokens are mapped to…

Machine Learning · Computer Science 2025-04-10 Golara Ahmadi Azar , Melika Emami , Alyson Fletcher , Sundeep Rangan

As new knowledge rapidly accumulates, language models (LMs) with pretrained knowledge quickly become obsolete. A common approach to updating LMs is fine-tuning them directly on new knowledge. However, recent studies have shown that…

Computation and Language · Computer Science 2025-02-28 Howard Chen , Jiayi Geng , Adithya Bhaskar , Dan Friedman , Danqi Chen

There are massive amounts of textual data residing in databases, valuable for many machine learning (ML) tasks. Since ML techniques depend on numerical input representations, word embeddings are increasingly utilized to convert symbolic…

Databases · Computer Science 2020-01-23 Michael Günther , Maik Thiele , Wolfgang Lehner

Large language models exhibit sophisticated capabilities, yet understanding how they work internally remains a central challenge. A fundamental obstacle is that training selects for behavior, not circuitry, so many weight configurations can…

Machine Learning · Computer Science 2026-02-27 Joshua S. Schiffman

Fact knowledge memorization is crucial for Large Language Models (LLM) to generate factual and reliable responses. However, the behaviors of LLM fact memorization remain under-explored. In this paper, we analyze the scaling laws for LLM's…

Computation and Language · Computer Science 2024-06-25 Xingyu Lu , Xiaonan Li , Qinyuan Cheng , Kai Ding , Xuanjing Huang , Xipeng Qiu