paper:arxiv-2502-00873Language models use trigonometry to do addition
Related work— refs + corpus + external arXiv
Cited / in-corpus / arXiv badges show which signals surfaced each row. Multi-source rows weighted higher.
- ≈ 70%
- Learning to Model the World with LanguageYuqing Du, Olivia Watkins, Danijar Hafner, Pieter Abbeel, Dan Klein, Anca Dragan Jessy Lin2024≈ 70%
- ≈ 70%
- Exploring Geographic Relative Space in Large Language Models through Activation PatchingRahul Baiju, Stefano Mizzaro, Kevin Roitero Stef De Sabbata2026≈ 70%
- ≈ 70%
- Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated ProbabilitiesSathvik Nair and Colin Phillips2026≈ 69%
- Evaluating Language Model Agency through NegotiationsVeniamin Veselovsky, Martin Josifoski, Maxime Peyrard, Antoine Bosselut, Michal Kosinski, Robert West Tim R. Davidson2026≈ 69%
- A Survey of Large Language ModelsKun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Zhipeng Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, Peiyu Liu, Jian-Yun Nie and Ji-Rong Wen Wayne Xin Zhao2026≈ 69%
- Advancing the Scientific Method with Large Language Models: From Hypothesis to DiscoverySumeer A. Khan, Adnan Mahmud, Huck Yang, Alexander Lavin, Michael Levin, Jeremy Frey, Jared Dunnmon, James Evans, Alan Bundy, Saso Dzeroski, Jesper Tegner, Hector Zenil Yanbo Zhang2025≈ 69%
- Multi-Agent Language Models: Advancing Cooperation, Coordination, and AdaptationArjun Vaithilingam Sudhakar2025≈ 68%
- Social Meaning in Large Language Models: Structure, Magnitude, and Pragmatic PromptingRoland M\"uhlenbernd2026≈ 68%
- The use of GPT-4o and Other Large Language Models for the Improvement and Design of Self-Assessment Scales for Measurement of Interpersonal Communication SkillsGoran Buba\v{s}2024≈ 68%
- Understanding Counting Mechanisms in Large Language and Vision-Language ModelsAmirmohammad Izadi, Fatemeh Askari, Mobin Bagherian, Sadegh Mohammadian, Mohammad Izadi, Mahdieh Soleymani Baghshah Hosein Hasani2026≈ 68%
- Language Models Meet World Models: Embodied Experiences Enhance Language ModelsTianhua Tao, Yi Gu, Tianmin Shu, Zirui Wang, Zichao Yang, Zhiting Hu Jiannan Xiang2023≈ 68%
- ≈ 68%
- ≈ 66%
- ≈ 66%
- ≈ 66%
- Evaluating Language Model Character Traitsin corpus2024≈ 65%
- ≈ 65%
- ≈ 64%
- ≈ 64%
- Interpreting Language Model Parametersin corpus2026≈ 64%
- ≈ 63%
- ≈ 63%
- ≈ 63%
- ≈ 63%
- Model Alignment Searchin corpus2025≈ 63%
- The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasetsin corpus2023≈ 63%
Similar preprints — Semantic Scholar
Cited by (7)
- From Memorization to Reasoning in the Spectrum of Loss Curvature
- Do Sparse Autoencoders Capture Concept Manifolds?
Sparse autoencoders trained on Llama3.1-8B (layer 19, 500M Pile tokens, 25 SAEs across Standard, JumpReLU, TopK, BatchTopK and Matryoshka architectures at VE > 0.85) do not capture continuous concept
- Structuring Sparsity: Block-Sparse Featurizers Capture Visual Concept Manifolds
Lifting the featurizer's atom from a direction to a block of directions recovers concept manifolds that SAEs shatter, and block sparsity is the MAP-matched prior for a representation that is a sparse
- When Models Manipulate Manifolds: The Geometry of a Counting Task
Claude 3.5 Haiku's linebreaking in fixed-width text is implemented by manipulating low-dimensional, rippled 1-D feature manifolds: character count, line width, characters remaining and next-token leng
- The Non-Linear Representation Dilemma: Is Causal Abstraction Enough for Mechanistic Interpretability?
Under arbitrarily powerful alignment maps, causal abstraction becomes vacuous: any neural network can be perfectly mapped to any algorithm, a result proven formally in Theorem 1 under five mild assump
- Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
Manifold steering — intervening on model activations along paths constrained to lie on a learned activation manifold M_h rather than along Euclidean linear directions — produces behavioral trajectorie
- Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts
Llama-3.1-8B solves cyclic arithmetic (e.g., "what month is six months after August?") not by performing modular addition in the period of the cyclic concept (12 for months, 7 for days of the week) as