{"total":216,"rows":[{"id":"2311.03658","title":"The Linear Representation Hypothesis and the Geometry of Large Language Models","created":"2023-11-07","cited_by_count":716,"area":null},{"id":"2304.14767","title":"Dissecting Recall of Factual Associations in Auto-Regressive Language Models","created":"2023-04-28","cited_by_count":528,"area":null},{"id":"2305.01610","title":"Finding Neurons in a Haystack: Case Studies with Sparse Probing","created":"2023-05-02","cited_by_count":334,"area":null},{"id":"2103.11790","title":"Large Pre-trained Language Models Contain Human-like Biases of What is Right and Wrong to Do","created":"2021-03-08","cited_by_count":178,"area":null},{"id":"2406.01506","title":"The Geometry of Categorical and Hierarchical Concepts in Large Language Models","created":"2024-06-03","cited_by_count":159,"area":null},{"id":"2108.04840","title":"Post-hoc Interpretability for Neural NLP: A Survey","created":"2021-08-10","cited_by_count":156,"area":null},{"id":"2109.04404","title":"All Bark and No Bite: Rogue Dimensions in Transformer Language Models Obscure Representational Quality","created":"2021-09-09","cited_by_count":156,"area":null},{"id":"2209.11326","title":"Towards Faithful Model Explanation in NLP: A Survey","created":"2022-09-22","cited_by_count":130,"area":null},{"id":"2207.13243","title":"Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networks","created":"2022-07-27","cited_by_count":129,"area":null},{"id":"2102.08473","title":"COCO-LM: Correcting and Contrasting Text Sequences for Language Model Pretraining","created":"2021-02-16","cited_by_count":126,"area":null},{"id":"2109.06129","title":"Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color","created":"2021-09-13","cited_by_count":119,"area":null},{"id":"2103.15949","title":"Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factors","created":"2021-03-29","cited_by_count":108,"area":null},{"id":"2103.08907","title":"BBAM: Bounding Box Attribution Map for Weakly Supervised Semantic and Instance Segmentation","created":"2021-03-16","cited_by_count":105,"area":null},{"id":"2305.08809","title":"Interpretability at Scale: Identifying Causal Mechanisms in Alpaca","created":"2023-05-15","cited_by_count":103,"area":null},{"id":"2104.06976","title":"Pose Recognition with Cascade Transformers","created":"2021-04-14","cited_by_count":99,"area":null},{"id":"2303.11504","title":"Language Model Behavior: A Comprehensive Survey","created":"2023-03-20","cited_by_count":92,"area":null},{"id":"2403.03867","title":"On the Origins of Linear Representations in Large Language Models","created":"2024-03-06","cited_by_count":79,"area":null},{"id":"2104.14795","title":"Mitigating Political Bias in Language Models Through Reinforced Calibration","created":"2021-04-30","cited_by_count":73,"area":null},{"id":"2101.00403","title":"Superbizarre Is Not Superb: Derivational Morphology Improves BERT's Interpretation of Complex Words","created":"2021-01-02","cited_by_count":63,"area":null},{"id":"2107.00758","title":"The Spotlight: A General Method for Discovering Systematic Errors in Deep Learning Models","created":"2021-07-01","cited_by_count":63,"area":null},{"id":"2205.07237","title":"Discovering Latent Concepts Learned in BERT","created":"2022-05-15","cited_by_count":56,"area":null},{"id":"2108.07344","title":"IsoScore: Measuring the Uniformity of Embedding Space Utilization","created":"2021-08-16","cited_by_count":53,"area":null},{"id":"2109.07152","title":"Incorporating Residual and Normalization Layers into Analysis of Masked Language Models","created":"2021-09-15","cited_by_count":51,"area":null},{"id":"2104.03869","title":"Probing BERT in Hyperbolic Spaces","created":"2021-04-08","cited_by_count":50,"area":null},{"id":"2305.03210","title":"AttentionViz: A Global View of Transformer Attention","created":"2023-05-04","cited_by_count":50,"area":null},{"id":"2112.11668","title":"How Should Pre-Trained Language Models Be Fine-Tuned Towards Adversarial Robustness?","created":"2021-12-22","cited_by_count":49,"area":null},{"id":"2205.05092","title":"Problems with Cosine as a Measure of Embedding Similarity for High Frequency Words","created":"2022-05-10","cited_by_count":49,"area":null},{"id":"2303.08117","title":"Do Transformers Parse while Predicting the Masked Word?","created":"2023-03-14","cited_by_count":48,"area":null},{"id":"2310.20320","title":"Theory of Mind in Large Language Models: Examining Performance of 11 State-of-the-Art models vs. Children Aged 7-10 on Advanced Tests","created":"2023-10-31","cited_by_count":48,"area":null},{"id":"2307.14023","title":"Are Transformers with One Layer Self-Attention Using Low-Rank Weight Matrices Universal Approximators?","created":"2023-07-26","cited_by_count":47,"area":null},{"id":"2601.04480","title":"When Models Manipulate Manifolds: The Geometry of a Counting Task","created":"2026-01-08","cited_by_count":46,"area":null},{"id":"2205.03286","title":"GlobEnc: Quantifying Global Token Attribution by Incorporating the Whole Encoder Layer in Transformers","created":"2022-05-06","cited_by_count":44,"area":null},{"id":"2301.12971","title":"Quantifying Context Mixing in Transformers","created":"2023-01-30","cited_by_count":43,"area":null},{"id":"2205.01287","title":"SemAttack: Natural Textual Attacks via Different Semantic Spaces","created":"2022-05-03","cited_by_count":42,"area":null},{"id":"2105.07109","title":"The Low-Dimensional Linear Geometry of Contextualized Word Representations","created":"2021-05-15","cited_by_count":40,"area":null},{"id":"2104.14694","title":"Let's Play Mono-Poly: BERT Can Reveal Words' Polysemy Level and Partitionability into Senses","created":"2021-04-29","cited_by_count":39,"area":null},{"id":"2105.14002","title":"What if This Modified That? Syntactic Interventions via Counterfactual Embeddings","created":"2021-05-28","cited_by_count":39,"area":null},{"id":"2206.13289","title":"Analyzing Encoded Concepts in Transformer Language Models","created":"2022-06-27","cited_by_count":38,"area":null},{"id":"2109.08359","title":"Distilling Linguistic Context for Language Model Compression","created":"2021-09-17","cited_by_count":33,"area":"distill"},{"id":"2105.10185","title":"A Non-Linear Structural Probe","created":"2021-05-21","cited_by_count":31,"area":null},{"id":"2206.14017","title":"Proton: Probing Schema Linking Information from Pre-trained Language Models for Text-to-SQL Parsing","created":"2022-06-28","cited_by_count":30,"area":null},{"id":"2309.07418","title":"A Fast Optimization View: Reformulating Single Layer Attention in LLM Based on Tensor and SVM Trick, and Solving It in Matrix Multiplication Time","created":"2023-09-14","cited_by_count":30,"area":null},{"id":"2104.01477","title":"Exploring the Role of BERT Token Representations to Explain Sentence Probing Results","created":"2021-04-03","cited_by_count":29,"area":null},{"id":"2106.01183","title":"A Cluster-based Approach for Improving Isotropy in Contextual Embedding Space","created":"2021-06-02","cited_by_count":27,"area":null},{"id":"2410.04010","title":"Hyperbolic Fine-Tuning for Large Language Models","created":"2024-10-05","cited_by_count":27,"area":null},{"id":"2410.09087","title":"Mechanistic?","created":"2024-10-07","cited_by_count":25,"area":"mechinterp"},{"id":"2104.05274","title":"Learning to Remove: Towards Isotropic Pre-trained BERT Embedding","created":"2021-04-12","cited_by_count":23,"area":null},{"id":"2109.04740","title":"How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy","created":"2021-09-10","cited_by_count":23,"area":null},{"id":"2206.04727","title":"STNDT: Modeling Neural Population Activity with a Spatiotemporal Transformer","created":"2022-06-09","cited_by_count":23,"area":null},{"id":"2308.10502","title":"GradientCoin: A Peer-to-Peer Decentralized Large Language Models","created":"2023-08-21","cited_by_count":23,"area":null},{"id":"2407.14662","title":"Relational Composition in Neural Networks: A Survey and Call to Action","created":"2024-07-19","cited_by_count":23,"area":null},{"id":"2201.06227","title":"Egeria: Efficient DNN Training with Knowledge-Guided Layer Freezing","created":"2022-01-17","cited_by_count":22,"area":null},{"id":"2402.12011","title":"A Systematic Comparison of Contextualized Word Embeddings for Lexical Semantic Change","created":"2024-02-19","cited_by_count":21,"area":null},{"id":"2410.17770","title":"Small Singular Values Matter: A Random Matrix Analysis of Transformer Models","created":"2024-10-23","cited_by_count":21,"area":null},{"id":"2102.06820","title":"Characterizing English Variation across Social Media Communities with BERT","created":"2021-02-12","cited_by_count":20,"area":null},{"id":"2104.08465","title":"Frequency-based Distortions in Contextualized Word Embeddings","created":"2021-04-17","cited_by_count":20,"area":null},{"id":"2310.19671","title":"Large Language Models: The Need for Nuance in Current Debates and a Pragmatic Perspective on Understanding","created":"2023-10-30","cited_by_count":20,"area":null},{"id":"2507.17787","title":"Hyperbolic Deep Learning for Foundation Models: A Survey","created":"2025-07-23","cited_by_count":20,"area":"mllm"},{"id":"2110.10470","title":"Interpreting Deep Learning Models in Natural Language Processing: A Review","created":"2021-10-20","cited_by_count":19,"area":null},{"id":"2301.11975","title":"Byte Pair Encoding for Symbolic Music","created":"2023-01-27","cited_by_count":19,"area":null},{"id":"2109.07971","title":"Do Language Models Know the Way to Rome?","created":"2021-09-16","cited_by_count":18,"area":null},{"id":"2109.09237","title":"MirrorWiC: On Eliciting Word-in-Context Representations from Pretrained Language Models","created":"2021-09-19","cited_by_count":18,"area":null},{"id":"2310.04861","title":"Uncovering hidden geometry in Transformers via disentangling position and context","created":"2023-10-07","cited_by_count":18,"area":null},{"id":"2105.12449","title":"LMMS Reloaded: Transformer-based Sense Embeddings for Disambiguation and Beyond","created":"2021-05-26","cited_by_count":17,"area":null},{"id":"2308.05576","title":"Do Language Models' Words Refer?","created":"2023-08-10","cited_by_count":17,"area":null},{"id":"2105.13471","title":"Inspecting the concept knowledge graph encoded by modern language models","created":"2021-05-27","cited_by_count":15,"area":null},{"id":"2310.17715","title":"Outlier Dimensions Encode Task-Specific Knowledge","created":"2023-10-26","cited_by_count":15,"area":null},{"id":"2409.03662","title":"The representation landscape of few-shot learning and fine-tuning in large language models","created":"2024-09-05","cited_by_count":15,"area":null},{"id":"2104.07578","title":"Syntactic Perturbations Reveal Representational Correlates of Hierarchical Phrase Structure in Pretrained Language Models","created":"2021-04-15","cited_by_count":14,"area":null},{"id":"2111.08546","title":"Interpreting Language Models Through Knowledge Graph Extraction","created":"2021-11-16","cited_by_count":14,"area":null},{"id":"2204.06201","title":"Probing for Constituency Structure in Neural Language Models","created":"2022-04-13","cited_by_count":14,"area":null},{"id":"2309.00254","title":"Why do universal adversarial attacks work on large language models?: Geometry might be the answer","created":"2023-09-01","cited_by_count":14,"area":null},{"id":"2104.10833","title":"Low Anisotropy Sense Retrofitting (LASeR) : Towards Isotropic and Sense Enriched Representations","created":"2021-04-22","cited_by_count":13,"area":null},{"id":"2302.02178","title":"Construction Grammar Provides Unique Insight into Neural Language Models","created":"2023-02-04","cited_by_count":13,"area":null},{"id":"2310.09925","title":"Homophone Disambiguation Reveals Patterns of Context Mixing in Speech Transformers","created":"2023-10-15","cited_by_count":13,"area":null},{"id":"2404.12545","title":"Latent Concept-based Explanation of NLP Models","created":"2024-04-18","cited_by_count":13,"area":null},{"id":"2405.14535","title":"Exploring Alignment in Shared Cross-lingual Spaces","created":"2024-05-23","cited_by_count":13,"area":null},{"id":"2204.09722","title":"When Does Syntax Mediate Neural Language Model Performance? Evidence from Dropout Probes","created":"2022-04-20","cited_by_count":12,"area":null},{"id":"2305.02469","title":"The System Model and the User Model: Exploring AI Dashboard Design","created":"2023-05-04","cited_by_count":12,"area":null},{"id":"2310.11923","title":"Investigating semantic subspaces of Transformer sentence embeddings through linear structural probing","created":"2023-10-18","cited_by_count":12,"area":null},{"id":"2504.08912","title":"HyperCore: The Core Framework for Building Hyperbolic Foundation Models with Comprehensive Modules","created":"2025-04-11","cited_by_count":12,"area":null},{"id":"2602.02464","title":"From Directions to Regions: Decomposing Activations in Language Models via Local Geometry","created":"2026-02-02","cited_by_count":12,"area":null},{"id":"2405.05386","title":"Interpretability Needs a New Paradigm","created":"2024-05-08","cited_by_count":11,"area":null},{"id":"2502.05390","title":"Learning Task Representations from In-Context Learning","created":"2025-02-08","cited_by_count":11,"area":null},{"id":"2109.02508","title":"Uniform Manifold Approximation and Projection (UMAP) and its Variants: Tutorial and Survey","created":"2021-08-25","cited_by_count":10,"area":null},{"id":"2302.02060","title":"Representation Deficiency in Masked Language Modeling","created":"2023-02-04","cited_by_count":10,"area":null},{"id":"2309.12934","title":"TOPFORMER: Topology-Aware Authorship Attribution of Deepfake Texts with Diverse Writing Styles","created":"2023-09-22","cited_by_count":10,"area":null},{"id":"2105.02629","title":"Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach","created":"2021-05-06","cited_by_count":9,"area":null},{"id":"2106.13553","title":"Exploring the Representation of Word Meanings in Context: A Case Study on Homonymy and Synonymy","created":"2021-06-25","cited_by_count":9,"area":null},{"id":"2109.11491","title":"Putting Words in BERT's Mouth: Navigating Contextualized Vector Spaces with Pseudowords","created":"2021-09-23","cited_by_count":9,"area":null},{"id":"2207.08466","title":"What does Transformer learn about source code?","created":"2022-07-18","cited_by_count":9,"area":null},{"id":"2210.03575","title":"Are Representations Built from the Ground Up? An Empirical Examination of Local Composition in Language Models","created":"2022-10-07","cited_by_count":9,"area":null},{"id":"2301.04704","title":"SensePOLAR: Word sense aware interpretability for pre-trained contextual word embeddings","created":"2023-01-11","cited_by_count":9,"area":null},{"id":"2401.15393","title":"Semantics of Multiword Expressions in Transformer-Based Models: A Survey","created":"2024-01-27","cited_by_count":9,"area":null},{"id":"2109.07319","title":"InceptionXML: A Lightweight Framework with Synchronized Negative Sampling for Short Text Extreme Classification","created":"2021-09-13","cited_by_count":8,"area":null},{"id":"2306.06205","title":"Morphosyntactic probing of multilingual BERT models","created":"2023-06-09","cited_by_count":8,"area":null},{"id":"2401.14040","title":"(Chat)GPT v BERT: Dawn of Justice for Semantic Change Detection","created":"2024-01-25","cited_by_count":8,"area":null},{"id":"2402.14880","title":"Automatic Histograms: Leveraging Language Models for Text Dataset Exploration","created":"2024-02-21","cited_by_count":8,"area":null},{"id":"2103.01819","title":"The Rediscovery Hypothesis: Language Models Need to Meet Linguistics","created":"2021-03-02","cited_by_count":7,"area":null},{"id":"2206.01512","title":"Latent Topology Induction for Understanding Contextualized Representations","created":"2022-06-03","cited_by_count":7,"area":null}],"next_cursor":"100"}