Language models

…CellOLMo, a collaboration using open language models with single-cell brain data. The researcher will develop multimodal models that combine gene-expression data and text, evaluate biological language grounding on the SEA-AD Alzheimer's disease atlas, build donor- and brain-region-level representations, and release resulting models, code, data, and research artifacts openly.

…computer vision, machine learning and language models. Responsibilities include research publication, collaboration and support for junior researchers. The Tokyo appointment is renewed annually, subject to evaluation and project continuation, through March 2029 at the latest. Applications are accepted until the positions are filled; the official vacancy e

…memory and knowledge management for language models and agents; reusable skills from agent execution histories; algorithms that reduce computing and memory requirements; and AI operating under device, edge and communication constraints. The researcher will formulate problems, implement methods, evaluate reliability and efficiency, publish papers and release software. The Tokyo position uses renewable annual contracts, with maximum terms of seven years for research scientists and five for po

Work with Ai2’s AllenNLP team on language models, natural-language processing and the open-model ecosystem. The programme prepares researchers for doctoral study through mentored projects, coauthored publications, and access to Ai2’s data, infrastructure and research tools. Its scientific agenda includes improving NLP performance and accountability and developing methods to evaluate and understand language systems. The role is based at Ai2’s Seattle offices, with on-site requirements determined by the team. The advertised compensation is USD 113,300. Candidates already pursuing a doctorate ma

…can be explained, suggesting that language models may combine many tractable, largely interpolative mechanisms. It then distinguishes that scientific account from a more speculative argument about the role of industrial computation and human labour, and what these imply for claims of broadly extrapolative intelligence.

…and dialogue remain open problems for language models. The conversation examines early abstraction of verb categories in small transformers, how distributed representations shape learning, and whether language alone can support meaningful representations. It also considers language in world models, diffusion language models, and representation finetuning (ReFT), which steers frozen models through their hidden states. The final discussion asks where knowledge resi

Chandra Sripada presents his argument that language models and humans share meaningful features of cognition. Drawing on cognitive science, he and Sean Carroll examine what behavioral similarities can tell us about underlying mechanisms and how to assess the relationship between artificial and human intelligence.

…talk uses Craik's account of mental models—supporting alternatives, prediction and transfer—to frame empirical tests. Spatial reasoning in language models and motion forecasting in visual models illustrate ways to probe those abilities, along with the limits of inferring world models from systems trained on human-produced representations.

Michael Riegler discusses current uses of AI and language models, possible future developments and the concerns accompanying their spread. Conversation in Norwegian.

Adam Brown reviews recent progress in training language models to perform scientific work and reasoning. The talk considers how the rapid growth of these capabilities could change theoretical-physics research, while distinguishing present achievements from speculation about the consequences of continuing progress.

Alexander Meiburg examines how language models and interactive theorem provers are changing AI-assisted mathematics. Models can propose creative natural-language arguments but also make subtle errors; formal proof systems provide grounding. The talk discusses model development, strengths and weaknesses, and the differences involved in extending formal assistance to physics, where establishing suitable definitions can be a larger obstacle than proving statements.

…those ideas to verbal thought, language models and the experience of establishing an independent research foundation.

Modern artificial intelligence (AI) systems are powered by foundation models. This paper presents a new set of foundation models, called Llama 3. It is a herd of language models that natively support multilinguality, coding, reasoning, and tool usage. Our largest model is a dense Transformer with 405B parameters and a context window of up to 128K tokens. This paper presents an extensive empirical evaluation of Llama 3. We find that Llama 3 delivers comparable quality to leading language models such as GPT-4 on a plethora of tasks. We publicly release Llama 3, including pre-trained and post-tra