LoRA, PEFT, instruction tuning and domain adaptation — adapting a pretrained model without paying to train one.
arXiv:2608.31073v1 Announce Type: new Abstract: Whole-heart segmentation (WHS) in computed tomography (CT) and magnetic resonance imaging (MRI) is affected by acquisition shifts and heterogeneous car...
arXiv:2608.24621v2 Announce Type: replace Abstract: Can language models be trusted in safety- critical operations? In such settings, strong per- formance on semantic metrics does not guaran- tee oper...
arXiv:2608.01291v2 Announce Type: replace Abstract: We present ArabicDialectSafety, a human-curated Arabic safety dataset of 25,071 prompts covering six Arabic varieties: Modern Standard Arabic, Syri...
The paper investigates how to choose language models (teachers) for generating multilingual synthetic data used to fine‑tune smaller student models. By evaluating 10 teacher models across six diverse languages and training 240 students, the authors find that teacher effectiveness is not driven by model size but by data qualities such as prompt diversity, length, and fluency, which explain most of the variance in student performance. Practical guidelines are offered, including matching teacher and student families and using translated prompts to improve outcomes for low‑resource languages.
arXiv:2608.30530v1 Announce Type: new Abstract: VLM-driven self-improvement of web code has a structural flaw: the model that proposes the repair is the model that judges it, and visual plausibility...
The paper introduces LLP, a Large Language Model–based generative framework for pricing second‑hand products on consumer‑to‑consumer platforms. LLP retrieves similar items to capture market dynamics, then uses LLMs to generate price suggestions, refined through supervised fine‑tuning and group relative policy optimization. A confidence‑based filter rejects unreliable predictions, and experiments show LLP outperforms prior methods, achieving higher static adoption rates when deployed on Xianyu.
arXiv:2605.25903v2 Announce Type: replace-cross Abstract: Activation verbalization explains hidden representations in natural language, but existing methods are mostly limited to self-explanation, wh...
arXiv:2604.12335v2 Announce Type: replace-cross Abstract: Training multimodal large language models (MLLMs) for video understanding requires large-scale annotated data spanning diverse tasks such as...
arXiv:2405.18540v3 Announce Type: replace-cross Abstract: Red-teaming, or identifying prompts that elicit harmful responses, is a critical step in ensuring the safe and responsible deployment of larg...
PaperGym is a framework that transforms each research paper into a training environment for AI research planning, using rubrics extracted from the paper’s method and experiments as a critic. It synthesizes research questions from the goal and background, and derives evaluation criteria from the method and experiments, reducing criterion leakage to 3.7%. Experiments with Qwen models show that training with PaperGym’s rubric improves benchmark performance by up to 5.6 points and outperforms existing datasets and fine‑tuning baselines.
arXiv:2605.27790v2 Announce Type: replace Abstract: Recent advances in large language models have accelerated open-vocabulary EEG-to-imagined-text decoding, where non-invasive neural activity recorde...
arXiv:2608.31096v1 Announce Type: cross Abstract: Class-incremental learning (CIL) requires a model to incrementally learn tasks that contain new classes without accessing earlier training data while...
CoVA‑SFT is a new large‑scale dataset comprising 51.9K samples and over 222K multimodal reasoning steps that teach models to interleave text and visual abstractions across five layout families and 17 complex tasks. It includes explicit rationale formulations, agentic renderings, and verification loops to help models build and maintain internal visual workspaces for purely textual reasoning problems. A companion benchmark, CoVA‑Bench, contains 1,700 held‑out test samples for reproducible evaluation, and models fine‑tuned on CoVA‑SFT outperform all interleaved CoT baselines by more than 2× on average, though they still lag behind strong text‑only CoT baselines.
The paper introduces Stratified Consistency Distillation, a fine‑tuning method to improve natural‑language to logical‑formula translations for neurosymbolic reasoning. It generates multiple candidate translations with a large LLM, clusters them by semantic equivalence, and selects pseudo‑labels using entropy‑based strategies—majority voting for low entropy, LLM-as‑Judge for medium entropy, and unification or abstention for high entropy. The selected pseudo‑labels are then used to fine‑tune a smaller model, yielding significant gains in Pass@K and a new Equivalent Logical Similarity metric.
arXiv:2608.29934v1 Announce Type: cross Abstract: KV-cache compression reduces LLM inference memory by evicting context tokens, but when the evicted tokens contain answer-bearing evidence, the model...
arXiv:2608.28967v1 Announce Type: cross Abstract: Electroencephalography (EEG)-based robotic control is commonly formulated as a direct classification problem, in which electrical neural signals are...
arXiv:2608.31013v1 Announce Type: new Abstract: Designing models that generalize effectively in low- to medium-data regimes remains a primary challenge in medical machine learning, particularly for p...
arXiv:2608.30720v1 Announce Type: new Abstract: Representational similarity is foundational to analyses of deep networks, yet distances between point-valued representations are not intrinsically tied...
arXiv:2605.28042v2 Announce Type: replace-cross Abstract: Modern large language models (LLMs) achieve state-of-the-art machine translation performance, but they do so as broad generalists largely tra...
arXiv:2605.27971v2 Announce Type: replace-cross Abstract: When large language models are fine-tuned to generate persona- or tone-conditioned responses, their output diversity is severely limited--a f...