{"total":63,"rows":[{"id":"2602.11451","title":"LoopFormer: Elastic-Depth Looped Transformers for Latent Reasoning via Shortcut Modulation","created":"2026-02-11","cited_by_count":61,"area":"lr"},{"id":"2502.01951","title":"On the Emergence of Position Bias in Transformers","created":"2025-02-04","cited_by_count":59,"area":null},{"id":"2407.09298","title":"Transformer Layers as Painters","created":"2024-07-12","cited_by_count":54,"area":null},{"id":"2504.15364","title":"KeyDiff: Key Similarity-Based KV Cache Eviction for Long-Context LLM Inference in Resource-Constrained Environments","created":"2025-04-21","cited_by_count":51,"area":"inference"},{"id":"2405.18781","title":"On the Role of Attention Masks and LayerNorm in Transformers","created":"2024-05-29","cited_by_count":48,"area":null},{"id":"2408.15417","title":"Implicit Geometry of Next-token Prediction: From Language Sparsity Patterns to Model Representations","created":"2024-08-27","cited_by_count":24,"area":null},{"id":"2510.11693","title":"Scaling Language-Centric Omnimodal Representation Learning","created":"2025-10-13","cited_by_count":22,"area":"mllm"},{"id":"2404.07647","title":"Why do small language models underperform? Studying Language Model Saturation via the Softmax Bottleneck","created":"2024-04-11","cited_by_count":21,"area":null},{"id":"2509.15714","title":"Once Upon a Time: Interactive Learning for Storytelling with Small Language Models","created":"2025-09-19","cited_by_count":11,"area":null},{"id":"2410.11462","title":"Mitigating Frequency Bias and Anisotropy in Language Model Pre-Training with Syntactic Smoothing","created":"2024-10-15","cited_by_count":10,"area":null},{"id":"2502.04469","title":"Ask and Remember: A Questions-Only Replay Strategy for Continual Visual Question Answering","created":"2025-02-06","cited_by_count":9,"area":null},{"id":"2508.17744","title":"Randomly Removing 50% of Dimensions in Text Embeddings has Minimal Impact on Retrieval and Classification Tasks","created":"2025-08-25","cited_by_count":8,"area":null},{"id":"2409.12737","title":"MEXMA: Token-level objectives improve sentence representations","created":"2024-09-19","cited_by_count":4,"area":null},{"id":"2503.02812","title":"Q-Filters: Leveraging QK Geometry for Efficient KV Cache Compression","created":"2025-03-04","cited_by_count":4,"area":"inference"},{"id":"2509.25359","title":"Geometric Metrics and LLMs: What They Measure and When They Work","created":"2025-09-29","cited_by_count":4,"area":null},{"id":"2510.24561","title":"LoRA-DA: Data-Aware Initialization for Low-Rank Adaptation via Asymptotic Analysis","created":"2025-10-28","cited_by_count":4,"area":null},{"id":"2407.17771","title":"Banyan: Improved Representation Learning with Explicit Structure","created":"2024-07-25","cited_by_count":3,"area":null},{"id":"2408.11868","title":"Improving embedding with contrastive fine-tuning on small datasets with expert-augmented scores","created":"2024-08-19","cited_by_count":3,"area":null},{"id":"2501.06562","title":"Discrete Speech Unit Extraction via Independent Component Analysis","created":"2025-01-11","cited_by_count":3,"area":null},{"id":"2502.16385","title":"Toward a Flexible Framework for Linear Representation Hypothesis Using Maximum Likelihood Estimation","created":"2025-02-22","cited_by_count":3,"area":null},{"id":"2601.18302","title":"Suppressing Final Layer Hidden State Jumps in Transformer Pretraining","created":"2026-01-26","cited_by_count":3,"area":null},{"id":"2602.20433","title":"Disentangling Geometry, Performance, and Training in Language Models","created":"2026-02-24","cited_by_count":3,"area":null},{"id":"2604.08764","title":"Revisiting Anisotropy in Language Transformers: The Geometry of Learning Dynamics","created":"2026-04-09","cited_by_count":3,"area":null},{"id":"2605.23033","title":"Uncovering the Latent Potential of Deep Intermediate Representations","created":"2026-05-21","cited_by_count":3,"area":null},{"id":"2504.12778","title":"Towards Lossless Token Pruning in Late-Interaction Retrieval Models","created":"2025-04-17","cited_by_count":2,"area":"inference"},{"id":"2605.18646","title":"Language-Switching Triggers Take a Latent Detour Through Language Models","created":"2026-05-18","cited_by_count":2,"area":"safety"},{"id":"2605.29782","title":"Hista and Numca: Estimate State Value Effectively for LLM Reinforcement Learning","created":"2026-05-28","cited_by_count":2,"area":null},{"id":"2606.03291","title":"Multilingual Unlearning in LLMs: Transfer, Dynamics, and Reversibility","created":"2026-06-02","cited_by_count":2,"area":"safety"},{"id":"2606.21514","title":"Towards Understanding the Power and Limits of the Muon Optimizer: A River-Valley Perspective","created":"2026-06-19","cited_by_count":2,"area":null},{"id":"2409.11253","title":"Norm of Mean Contextualized Embeddings Determines their Variance","created":"2024-09-17","cited_by_count":1,"area":null},{"id":"2503.09094","title":"Domain Adaptation for Japanese Sentence Embeddings with Contrastive Learning based on Synthetic Sentence Generation","created":"2025-03-12","cited_by_count":1,"area":null},{"id":"2506.11096","title":"Assessing the Impact of Anisotropy in Neural Representations of Speech: A Case Study on Keyword Spotting","created":"2025-06-06","cited_by_count":1,"area":null},{"id":"2509.20997","title":"Binary Autoencoder for Mechanistic Interpretability of Large Language Models","created":"2025-09-25","cited_by_count":1,"area":"mechinterp"},{"id":"2511.11041","title":"Correcting Mean Bias in Text Embeddings: A Refined Renormalization with Training-Free Improvements on MMTEB","created":"2025-11-14","cited_by_count":1,"area":null},{"id":"2602.13524","title":"Singular Vectors of Attention Heads Align with Features","created":"2026-02-13","cited_by_count":1,"area":"mechinterp"},{"id":"2603.13359","title":"From Refusal Tokens to Refusal Control: Discovering and Steering Category-Specific Refusal Directions","created":"2026-03-09","cited_by_count":1,"area":"alignment"},{"id":"2605.03196","title":"Geometric Deviation as an Unsupervised Pre-Generation Reliability Signal: Probing LLM Representations for Answerability","created":"2026-05-04","cited_by_count":1,"area":null},{"id":"2605.30596","title":"Improving Relative Representations with Learned Anchors and Whitened Inner Products","created":"2026-05-28","cited_by_count":1,"area":null},{"id":"2606.27378","title":"Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMs","created":"2026-05-07","cited_by_count":1,"area":"lr"},{"id":"2609.15758","title":"Sequential Adapter Stacking for Cross-Lingual Low-Resource ASR","created":"2026-09-14","cited_by_count":1,"area":null},{"id":"2410.12292","title":"How much do contextualized representations encode long-range context?","created":"2024-10-16","cited_by_count":0,"area":null},{"id":"2510.14925","title":"False Fixed Points: Kantian Feedback, Stable Miscalibration, and Representational Compression in LLMs","created":"2025-10-16","cited_by_count":0,"area":null},{"id":"2602.05035","title":"Capacity Constraints and the Multilingual Penalty for Lexical Disambiguation","created":"2026-02-04","cited_by_count":0,"area":null},{"id":"2602.17287","title":"Representation Collapse in Machine Translation Through the Lens of Angular Dispersion","created":"2026-02-19","cited_by_count":0,"area":null},{"id":"2603.04127","title":"Data-Aware Random Feature Kernel for Transformers","created":"2026-03-04","cited_by_count":0,"area":null},{"id":"2603.25385","title":"GlowQ: Group-Shared LOw-Rank Approximation for Quantized LLMs","created":"2026-03-26","cited_by_count":0,"area":null},{"id":"2604.16576","title":"On the Robustness of LLM-Based Dense Retrievers: A Systematic Analysis of Generalizability and Stability","created":"2026-04-17","cited_by_count":0,"area":null},{"id":"2604.21511","title":"From Tokens to Concepts: Leveraging SAE for SPLADE","created":"2026-04-23","cited_by_count":0,"area":null},{"id":"2605.09090","title":"Investigating Anisotropy in Visual Grounding under Controlled Counterfactual Perturbations","created":"2026-05-09","cited_by_count":0,"area":null},{"id":"2606.17037","title":"The Importance of Phase in Neural Representations: An Internal Oppenheim-Lim Test of Image Classifiers","created":"2026-06-15","cited_by_count":0,"area":null},{"id":"2606.19638","title":"MiqraBERT: Regression-Based Sentence-BERT Finetuning for Biblical Hebrew Parallel Detection","created":"2026-06-17","cited_by_count":0,"area":null},{"id":"2606.28343","title":"The Crowded Embedding Space: A Mean-Field Mechanism for Emergent Marginalization in Retrieval-Augmented Agents","created":"2026-06-01","cited_by_count":0,"area":"rag"},{"id":"2607.17019","title":"Regularize or Localize: When Training-Time KV-Cache Geometry Pays Under Quantization","created":"2026-07-19","cited_by_count":0,"area":"inference"},{"id":"2608.14819","title":"What Makes a Good Layer? Assessing the Layer-Wise Intrinsic Properties of Music Foundation Models","created":"2026-08-14","cited_by_count":0,"area":null},{"id":"2608.16269","title":"Domain-Agnostic Neural Topic Modeling with Contextual Token-Level Semantic Graph Representation","created":"2026-08-17","cited_by_count":0,"area":null},{"id":"2609.04819","title":"A Systematic Comparison of Multilingual Interpretability Methods Reveals Anisotropy-Driven Failures","created":"2026-09-04","cited_by_count":0,"area":null},{"id":"2609.06663","title":"ECOKV: Geometry-Aware KV Cache Eviction via Complementary Diversity Metrics","created":"2026-09-06","cited_by_count":0,"area":"mllm"},{"id":"2609.08692","title":"Global Divergence, Local Convergence: Representation Geometry in SSMs and Transformers","created":"2026-09-08","cited_by_count":0,"area":"ssm"},{"id":"2609.13198","title":"SPICE: Simple Polysemantic Feature Interpretation via Clustering-based Explanation","created":"2026-08-14","cited_by_count":0,"area":null},{"id":"2609.23314","title":"ValueDiff: Value-Geometric KV Cache Eviction for Sink-Suppressed LLMs","created":"2026-09-20","cited_by_count":0,"area":"inference"},{"id":"2609.32445","title":"Masking Frequent Tokens Sharpens Direct Preference Optimization","created":"2026-09-26","cited_by_count":0,"area":"alignment"},{"id":"2609.32792","title":"Understanding and Exploiting Anisotropy in Post-Training","created":"2026-09-26","cited_by_count":0,"area":null},{"id":"2610.06064","title":"TrustMI: Causally controlling how assistants trust their users","created":"2026-10-05","cited_by_count":0,"area":"safety"}],"next_cursor":null}