Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2608. 13517v1 Announce Type: cross Abstract: Current large language model development relies on massive, often non-permissible datasets, creating a high barrier for researchers committed to open-source and ethically sourced data.
arXiv:2608. 13545v1 Announce Type: cross Abstract: Modern language models are trained on heterogeneous web-scale text corpora.
arXiv:2608. 13459v1 Announce Type: cross Abstract: We address the use of large language models (LLMs) to help discover Isabelle proofs.
arXiv:2608. 13433v1 Announce Type: cross Abstract: Transformer-based language models are known to sometimes generalize to sequences longer than seen during training, but we lack a precise characterization of which tasks admit length generalization.
arXiv:2608. 12368v1 Announce Type: new Abstract: Agreement with human judgments is a common proxy for evaluating the alignment of large language models (LLMs).
arXiv:2608. 12395v1 Announce Type: new Abstract: We describe Research Assistant, an internal LLM-based system developed at AstraZeneca to help scientists and clinicians explore biomedical questions across a broad range of data sources.
arXiv:2608. 12675v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) is widely used to improve the performance of Large Language Models (LLMs) in answering user queries.
arXiv:2608. 13317v1 Announce Type: new Abstract: Large language model based multi-agent systems usually communicate in text, i.
arXiv:2608. 13420v1 Announce Type: new Abstract: Embodied intelligent virtual agents are expected to operate as persistent, adaptive, and context-aware entities within complex virtual and Metaverse worlds.
arXiv:2608. 12339v1 Announce Type: cross Abstract: Large Language models (LLMs) were found to be susceptible to a host of social, affective, and cognitive biases.
arXiv:2608. 12377v1 Announce Type: cross Abstract: Brains and large language models (LLMs) are fundamentally different memory systems, but they can be compared through shared functional questions: where memory-related information is represented, how partial cues recover broader associations, how new information is written or updated, and how memory-related states can be perturbed.
arXiv:2608. 12845v1 Announce Type: cross Abstract: Semantic ID (SID)-based generative recommendation has recently achieved remarkable success.
arXiv:2608. 13136v1 Announce Type: cross Abstract: With the rapid advancement of large language models (LLMs), research idea generation has attracted increasing attention.
arXiv:2608. 13258v1 Announce Type: cross Abstract: Self-referential prompting has been shown to reliably induce large language models to produce first-person reports resembling subjective experience, but no prior work measures how consistent these reports are across repeated, independent trials, or how that consistency compares to the model's behavior on other kinds of open-ended questions.
arXiv:2608. 13277v1 Announce Type: cross Abstract: We ask whether language-model pre-training can be decomposed into smaller, independently trainable jobs that can later be recomposed into a coherent larger model.
LLM agents increasingly maintain personal memory across sessions, but it can conflict. Preferences depend on context, behavior evolves, and sources can conflict. When a query lacks context, time, or s...
Release: llm-gemini 0. 33 It's been a while since the last llm-gemini release.
Sparse autoencoders (SAEs) are proposed to extract numerous features from large language model (LLM) representations, yet explaining these features still relies primarily on external observation. This reliance leads to superficial explanations inferred from observed model behavior and computational inefficiency from collecting such behavioral evidence at scale.
Speculative decoding losslessly accelerates autoregressive language models by verifying multiple draft tokens in parallel. Diffusion-based drafters further reduce proposal latency by predicting an entire token block in parallel, but their position-wise distributions are marginal rather than conditioned on tokens selected along each draft path.
Measuring training data influence consistently across language model pretraining is challenging. It is difficult to select downstream tasks or validation sets representative of a model's general capabilities, and reliance on task performance at intermediate checkpoints complicates comparisons across training.