large language models

GenAI for Research Support: A Case to DMPs Pre-Filling

A GenAI tool was developed that used a Large Language Model capable of answering typical questions for a DMP after receiving the original research proposal as input, and showed a very good capability in providing appropriate responses to most of the DMP questions.

Fine-Tuning Pre-trained Vision-Language Models in a Human-Annotation-Free Manner

Semantic Phase Theory: An Information-Theoretic Framework for Meaning and Context

This paper introduces Noetica Theory, a meaning-phase model in which semantic entities are represented as wave functions defined over context-dependent phase spaces. We define raw and normalized meaning wave functions, noetica energy, semantic entropy, semantic free energy, and semantic interference, and we formulate a...

functional-standard-atlas: attenuation-corrected, territory-resolved benchmarking of variant effect predictors against saturation genome editing

functional-standard-atlas: attenuation-corrected, territory-resolved benchmarking of variant effect predictors against saturation genome editing Software and frozen data underlying the manuscript "Measurement reliability bounds functional benchmarks and relocates where variant effect prediction fails" (N. Zhang). WHAT...

Parallel Dialogues: What LLM Inference Can (and Cannot) Teach Us About the Many-Worlds Interpretation — A Companion Note

A companion note to "Evidence for the Many-Worlds Interpretation of Quantum Mechanics via Large Language Model Architecture" (10.5281/zenodo.20533902). Revisits the structural analogy between parallel LLM inference and Everettian branching honestly: identifies what the analogy illustrates well (shared origin, isolated...

Parallel Dialogues: What LLM Inference Can (and Cannot) Teach Us About the Many-Worlds Interpretation — A Companion Note

A companion note to "Evidence for the Many-Worlds Interpretation of Quantum Mechanics via Large Language Model Architecture" (10.5281/zenodo.20533902). Revisits the structural analogy between parallel LLM inference and Everettian branching honestly: identifies what the analogy illustrates well (shared origin, isolated...

Dahono-4B: A Sovereign Foundation Model for Indonesian Legal Reasoning, Statutory Hierarchy Alignment, and Multimodal Document Intelligence

General-purpose Large Language Models (LLMs) predominantly trained on Anglo-American Common Law corpora demonstrate severe jurisdictional fragility when confronted with civil law traditions, frequently exhibiting statutory hallucination, citation of repealed legislation, and systematic disregard for statutory hierarchi...

Dahono-4B: A Sovereign Foundation Model for Indonesian Legal Reasoning, Statutory Hierarchy Alignment, and Multimodal Document Intelligence

General-purpose Large Language Models (LLMs) predominantly trained on Anglo-American Common Law corpora demonstrate severe jurisdictional fragility when confronted with civil law traditions, frequently exhibiting statutory hallucination, citation of repealed legislation, and systematic disregard for statutory hierarchi...

Information-Geometric Trajectory Routing

Large language model inference is usually optimized from the model inward: runtime systems predict future experts from hidden states, router scores, token statistics, or local activation correlations, while dynamic-depth methods decide which model layers can be skipped for the current input. This preprint proposes a di...

Carrier Current Sensing for Computation-Level Bias Detection in Large Language Models: A Methodological Manual

This is unedited work with Mistral Vibe. Quality/logic checked through multiple different instance reviews. Proceed with caution. This manual presents a novel methodology for detecting algorithmic bias in large language models (LLMs) by analyzing GPU power-line current fluctuations during inference. Unlike existing bia...

variant-fm-benchmark: ranking, calibration and evidence strength for splice-region variant effect predictors against a functional standard

Ranking is saturated, calibration is not, and neither changes the evidence strength — code and frozen analysis set Software and frozen data underlying the manuscript "Ranking is saturated, calibration is not, and neither changes the evidence strength: an independent benchmark of splice-region variant effect predictors...

Optimizing Few-Shot Learning in Pruned Large Language Models with Task-Specific Prompts

Few-shot learning enables large language models to efficiently perform tasks given only a limited number of labeled examples. However, training these models entirely from scratch requires substantial computational resources, making it challenging for many organizations to fully leverage their potential. This thesis exp...