LoRA, PEFT, instruction tuning and domain adaptation — adapting a pretrained model without paying to train one.
arXiv:2609.22896v1 Announce Type: new Abstract: Autonomous vehicles operating in open-world scenarios are inevitably confronted with previously unknown objects, such as exotic animals or loose cargo....
arXiv:2609.23990v1 Announce Type: new Abstract: Tracking neurite morphology over time helps characterize structural changes during neuronal development and deterioration, but long-term time-lapse ima...
arXiv:2609.22631v1 Announce Type: new Abstract: Accurate automated interpretation of electrocardio- grams (ECGs) is essential for early detection of cardiac condi- tions such as myocardial infarction...
arXiv:2609.23924v1 Announce Type: new Abstract: Pretrained EEG foundation models are increasingly proposed as general-purpose encoders for brain-computer interfaces, yet recent benchmarks disagree ab...
arXiv:2601.12538v2 Announce Type: replace-cross Abstract: Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) d...
arXiv:2609.23838v1 Announce Type: new Abstract: Greenhouse gas emissions estimates can be derived using inverse methods by combining atmospheric concentration observations with chemical transport mod...
arXiv:2604.13286v2 Announce Type: replace Abstract: Despite the widespread multilingual deployment of large language models, post-training pipelines remain predominantly English-centric, contributing...
arXiv:2609.23594v1 Announce Type: new Abstract: Orthogonality in a LoRA factor does not by itself specify what the composed update protects: the answer depends on the task-start state, the parameteri...
ZoAQ is an adaptive zeroth-order optimization method that reuses past function evaluations to inform both the next update and the decision to query further, eliminating the need for extra validation queries. The approach achieves significant query savings—43–48% fewer queries than fixed-budget baselines on synthetic tasks—and demonstrates strong performance in black-box adversarial attacks, achieving 100% success with 320 and 625 average queries on MNIST and CIFAR-10. In fine-tuning experiments across four settings, ZoAQ reduces forward evaluations by 43–46% while maintaining accuracy within ±0.018 to +0.010.
arXiv:2602.01654v2 Announce Type: replace Abstract: Steering vectors (SVs) offer a lightweight way to control large language models (LLMs) at inference time by shifting hidden activations, providing...
The paper investigates how conditioned floor plan generation models perform when applied to datasets from different regions, revealing significant performance drops due to domain shift. To address this, the authors create a large synthetic training set that enforces physical constraints while deliberately reducing architectural realism, and show that pre‑training on this data boosts zero‑shot cross‑domain performance and speeds up fine‑tuning in low‑data scenarios.
The paper investigates time‑incremental continued pretraining (CPT) of large language models using web‑scale data that overlaps across snapshots. Across six open‑weight models, CPT improves factual recall without catastrophic forgetting, while the cost is negligible and data quality outweighs quantity. The study identifies optimal learning rates, shows LoRA can match full CPT, and demonstrates that CPT gains transfer to fine‑tuned models.
The paper introduces a cross-dialect Named Entity Recognition (NER) framework for Bangla, leveraging the ANCHOLIK-NER dataset that covers five major regional dialects. Using a Leave-One-Dialect-Out Cross-Validation strategy, eight transformer-based models were evaluated, with Multilingual-E5 Large achieving the best performance (F1 up to 97.26% on Mymensingh, 82.38% on Chattogram). Local Interpretable Model-agnostic Explanations (LIME) revealed that the models rely mainly on the surface form of entity words rather than surrounding context, suggesting a direction for future improvement.
arXiv:2602.01161v2 Announce Type: replace Abstract: The global deployment of large language models (LLMs) has raised concerns about cultural misalignment, yet the linguistic properties of fine-tuning...
arXiv:2609.24657v1 Announce Type: cross Abstract: Language models can be adapted by changing the computations applied to individual tokens. Quantum circuits offer one such approach, but evaluating wi...
arXiv:2609.24903v1 Announce Type: new Abstract: Tone languages constitute over 50-70% of the world's languages, but the vast majority are low-resource, lacking the large transcribed corpora needed fo...
AdaptiveCDM is a modular framework for source‑free few‑shot domain adaptation in cell detection, enabling a pretrained model to adapt to new imaging domains using only a handful of labeled target images and no source data. It combines Resolution‑Aware Augmentation (RAug) to balance scarce, class‑imbalanced samples while preserving cellular morphology, and Category‑Aware Representation Learning (CARL) to strengthen class‑consistent proposals for better localization and classification. Experiments on M5 and Raabin‑WBC datasets show that AdaptiveCDM achieves competitive or superior mAP scores compared to state‑of‑the‑art methods under their respective supervision settings.
This thesis explores how to select and adapt NLP models for global health literature when annotated data and computational resources are scarce. It compares skip‑gram word2vec models trained on increasingly large specialized corpora with BioWordVec for semantic tag discovery, finding that larger coverage does not always yield more useful domain associations. The study also evaluates convolutional spaCy models versus a RoBERTa transformer for named entity recognition, noting a trade‑off between higher F1 scores and longer inference time, and investigates MiniLM few‑shot versus BART‑MNLI zero‑shot classification for multi‑label topic classification, highlighting practical constraints of inference cost. "whyItMatters":"The work provides empirical guidance on balancing model accuracy and resource demands for building knowledge systems in low‑resource global health settings."
The paper presents a smartphone‑compatible, retrieval‑augmented language model tailored to Bangladeshi statutory law. By distilling a 9‑billion‑parameter Gemma‑2 teacher into a 2‑billion‑parameter student using supervised fine‑tuning and QLoRA, the authors achieve significant gains in ROUGE‑L and BERTScore on an English benchmark while keeping the model lightweight (1.6 GB) and operable offline on a Pixel 6. The system retrieves from 36,029 statutory passages using a hybrid dense/BM25 approach, and cross‑lingual evaluation shows effective Bangla query handling against an English‑only corpus, with a practicing lawyer rating the responses highly in a pilot study.
Euston is an 8‑B parameter mathematical claim‑verification model that resists producing false derivations when presented with corrupted theorems. It was trained on 3,026 matched true/corrupted statement pairs generated by GraphSynth, a probabilistic factor‑graph generator, and fine‑tuned from DeepSeek‑R1‑8B using GRPO. On a balanced held‑out split, Euston’s balanced accuracy rose from 29.50 % to 63.75 %, and its discrimination gap improved from –0.5 % to +27.5 %, while maintaining comparable general mathematical ability and reducing response length and truncation rates.