cs updates on arXiv.org

Activation-Guided Neuron Intervention to Induce Alzheimer's-Related Computational Language Phenotypes in a Large Language Model

arXiv·

MDLMPE: Distribution Aware Positional Encoding for Masked Diffusion Language Models

arXiv·

string2string Studio: An Interactive, In-Browser Platform for String-to-String Algorithms

arXiv·

Probing Character-level Transformers for the Spanish L-shaped Morphome

arXiv·

ConlangBench: Exploring Language Knowledge and Learning in LLMs through Diverse Constructed Languages

arXiv·

Efficient Multilingual Neural Machine Translation via Corpus-Driven Vocabulary Pruning: An English-Arabic Case Study

arXiv·

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension

arXiv·

Predicting Multilingual Classification and Translation Performance of LLMs with Cross-Lingual Alignment $\unicode{x2013}$ Is English Enough?

arXiv·

Translation of Regular Expression with Lookahead into Finite State Automaton

arXiv·

Right Divisibility in Erasing Semi-Thue Systems: A Minimal View of Intruder Deduction

arXiv·

Certified Split Points for Parallel Lexing: Exact and Modulo Discarded Tokens

arXiv·

A positional $\mathbf{\Pi}^0_3$-complete objective

arXiv·

Beyond Accuracy: Auditing Spatial Provenance in Visual Token Pruning for OCR-Critical MLLM Inference

arXiv·

Noise-Robust Conditional Flow Matching: Generating Clean Samples from Noisy Datasets

arXiv·

Which Modality Decides? Counterfactual Modality Attribution for Multimodal LLMs

arXiv·

Video Models as Native 4D Renderers: World-Grounded Conditioning from Animated Mesh

arXiv·

IDraw: Artist Verification from Digital Drawing Images

arXiv·

K-space Gaussian Representation for Parallel MRI

arXiv·

AnchorKV: Anchor-Residual KV Cache Compression

Activation-Guided Neuron Intervention to Induce Alzheimer's-Related Computational Language Phenotypes in a Large Language Model

MDLMPE: Distribution Aware Positional Encoding for Masked Diffusion Language Models

string2string Studio: An Interactive, In-Browser Platform for String-to-String Algorithms

Probing Character-level Transformers for the Spanish L-shaped Morphome

ConlangBench: Exploring Language Knowledge and Learning in LLMs through Diverse Constructed Languages

Efficient Multilingual Neural Machine Translation via Corpus-Driven Vocabulary Pruning: An English-Arabic Case Study

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension

Predicting Multilingual Classification and Translation Performance of LLMs with Cross-Lingual Alignment $\unicode{x2013}$ Is English Enough?

Translation of Regular Expression with Lookahead into Finite State Automaton

Right Divisibility in Erasing Semi-Thue Systems: A Minimal View of Intruder Deduction

Certified Split Points for Parallel Lexing: Exact and Modulo Discarded Tokens

A positional $\mathbf{\Pi}^0_3$-complete objective

Beyond Accuracy: Auditing Spatial Provenance in Visual Token Pruning for OCR-Critical MLLM Inference

Noise-Robust Conditional Flow Matching: Generating Clean Samples from Noisy Datasets

Which Modality Decides? Counterfactual Modality Attribution for Multimodal LLMs

Video Models as Native 4D Renderers: World-Grounded Conditioning from Animated Mesh

IDraw: Artist Verification from Digital Drawing Images

K-space Gaussian Representation for Parallel MRI

Motion Beyond Morphology: Bootstrapping Cross-Category Motion Transfer from Abstract Motion Representations