1 citations · 1 across the 27 of their papers we have counts for
29 papers
Do LMs Know What Variables Do? Interpreting Code Representations via Variable Roles
Célian Vasson, Daniil Ognev, Kentaro Inui +1
Variable roles describe recurring behaviors such as accumulation, traversal, and retaining the best candidate encountered. We use Sajaniemi's independently defined taxonomy to stud…
Rethinking Contextualization by Reinterpreting Attention Head Channels
Hakaze Cho, Haolin Yang, Zhun Sun +3
Contextualization, the core operation of language modeling, transmits information across words to build sentence-specific word representations. Prior works mainly study contextuali…
SharedSAE: One Feature Dictionary Across Language Models
Daniil Ognev, Célian Vasson, Lijie Hu +2
Sparse autoencoders (SAEs) are widely used to interpret language model activations, but SAE training and latent labelling are typically repeated for every model. Here, we show that…
Relation Geometry in Semantic Space of Language Models
Zhihan Cao, Hiroaki Yamada, Simone Teufel +4
When it comes to generating vector representations of words, current language models are achieving high-quality results. However, what is not known is the extent to which knowledge…
Einstein World Models
Munachiso Samuel Nwadike, Zangir Iklassov, Ali Mekky +2
Does intelligence require the ability to reason about phenomena beyond direct experience? It is natural to suspect that some complex thought cannot be captured through language alo…
Language Models Compare Quantities Using Number-specific and Unit-specific Heuristics
Mutsumi Sasaki, Go kamoda, Ryosuke Takahashi +4
Quantities with measurement units, such as 110 cm and 1.2 m, require language models (LMs) to combine a numeral with a symbolic unit scale. Here, we study how LMs compare such quan…
Measuring AI Reasoning: A Guide for Researchers
Munachiso Samuel Nwadike, Zangir Iklassov, Kareem Ali +2
In this paper, we offer a guide for researchers on evaluating reasoning in language models, building the case that reasoning should be assessed through evidence of adaptive, multi-…
Cell-Based Representation of Relational Binding in Language Models
Qin Dai, Benjamin Heinzerling, Kentaro Inui
Understanding a discourse requires tracking entities and the relations that hold between them. While Large Language Models (LLMs) perform well on relational reasoning, the mechanis…
Linear Representations of Hierarchical Concepts in Language Models
Masaki Sakata, Benjamin Heinzerling, Takumi Ito +2
We investigate how and to what extent hierarchical relations (e.g., Japan Eastern Asia Asia) are encoded in the internal representations of language models. Bui…
Nodes Are Early, Edges Are Late: Probing Diagram Representations in Large Vision-Language Models
Haruto Yoshida, Keito Kudo, Yoichi Aoki +4
Large vision-language models (LVLMs) demonstrate strong performance on diagram understanding benchmarks, yet they still struggle with understanding relationships between elements,…
Can Language Models Handle a Non-Gregorian Calendar? The Case of the Japanese wareki
Mutsumi Sasaki, Go Kamoda, Ryosuke Takahashi +4
Temporal reasoning and knowledge are essential capabilities for language models (LMs). While much prior work has analyzed and improved temporal reasoning in LMs, most studies have…
On Entity Identification in Language Models
Masaki Sakata, Benjamin Heinzerling, Sho Yokoi +2
We analyze the extent to which internal representations of language models (LMs) identify and distinguish mentions of named entities, focusing on the many-to-many correspondence be…
TopK Language Models
Ryosuke Takahashi, Tatsuro Inaba, Kentaro Inui +1
Sparse autoencoders (SAEs) have become an important tool for analyzing and interpreting the activation space of transformer-based language models (LMs). However, SAEs suffer severa…
Do LLMs Need to Think in One Language? Correlation between Latent Language and Task Performance
Shintaro Ozaki, Tatsuya Hiraoka, Hiroto Otake +8
Large Language Models (LLMs) are known to process information using a proficient internal language consistently, referred to as latent language, which may differ from the input or…
Understanding Fact Recall in Language Models: Why Two-Stage Training Encourages Memorization but Mixed Training Teaches Knowledge
Ying Zhang, Benjamin Heinzerling, Dongyuan Li +1
While fine-tuning is the standard for injecting factual knowledge into large language models (LLMs), the mechanisms enabling reliable fact recall via unseen queries remain poorly u…
SPIRIT: Patching Speech Language Models against Jailbreak Attacks
Amirbek Djanibekov, Nurdaulet Mukhituly, Kentaro Inui +2
Speech Language Models (SLMs) enable natural interactions via spoken instructions, which more effectively capture user intent by detecting nuances in speech. The richer speech sign…
How a Bilingual LM Becomes Bilingual: Tracing Internal Representations with Sparse Autoencoders
Tatsuro Inaba, Go Kamoda, Kentaro Inui +5
This study explores how bilingual language models develop complex internal representations. We employ sparse autoencoders to analyze internal representations of bilingual language…
Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference
Go Kamoda, Benjamin Heinzerling, Tatsuro Inaba +3
According to the stages-of-inference hypothesis, early layers of language models map their subword-tokenized input, which does not necessarily correspond to a linguistically meanin…
FinchGPT: a Transformer based language model for birdsong analysis
Kosei Kobayashi, Kosuke Matsuzaki, Masaya Taniguchi +3
The long-range dependencies among the tokens, which originate from hierarchical structures, are a defining hallmark of human language. However, whether similar dependencies exist w…
RECALL: Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles
Munachiso Nwadike, Zangir Iklassov, Toluwani Aremu +6
We introduce the concept of the self-referencing causal cycle (abbreviated RECALL) - a mechanism that enables large language models (LLMs) to bypass the limitations of unidirection…
Representational Analysis of Binding in Language Models
Qin Dai, Benjamin Heinzerling, Kentaro Inui
Entity tracking is essential for complex reasoning. To perform in-context entity tracking, language models (LMs) must bind an entity to its attribute (e.g., bind a container to its…
Repetition Neurons: How Do Language Models Produce Repetitions?
Tatsuya Hiraoka, Kentaro Inui
This paper introduces repetition neurons, regarded as skill neurons responsible for the repetition problem in text generation tasks. These neurons are progressively activated more…
The Geometry of Numerical Reasoning: Language Models Compare Numeric Properties in Linear Subspaces
Ahmed Oumar El-Shangiti, Tatsuya Hiraoka, Hilal AlQuabeh +2
This paper investigates whether large language models (LLMs) utilize numerical attributes encoded in a low-dimensional subspace of the embedding space when answering questions invo…
The Curse of Popularity: Popular Entities have Catastrophic Side Effects when Deleting Knowledge from Language Models
Ryosuke Takahashi, Go Kamoda, Benjamin Heinzerling +2
Language models (LMs) encode world knowledge in their internal parameters through training. However, LMs may learn personal and confidential information from the training data, lea…