4Berkeley AI Research (BAIR) Blog·
BAIR Lab 2026 PhD Graduate Showcase: Placements at OpenAI, Physical Intelligence, Mistral, and Academia
Berkeley Artificial Intelligence Research (BAIR) Lab published its 2026 graduate showcase, highlighting PhD completions across LLMs, robotics, AI safety, computer vision, and human-AI interaction. Notable placements include a graduate joining OpenAI as Member of Technical Staff (LLM reasoning), one joining Physical Intelligence (generalist vision/robotics), one joining Mistral AI as AI Scientist, and one becoming an Assistant Professor at UCLA. The cohort's research themes span test-time vs. pretraining scaling tradeoffs, LLM fairness and calibration, dexterous manipulation, and generative modeling for proteins.
Related guides (4)
Related events (8)
Mistral AI Highlights Physics AI Research Portfolio Following Emmi AI Acquisition
Mistral AI has published a research overview showcasing Physics AI work stemming from its acquisition of Emmi AI, targeting industrial engineering domains including aerospace, automotive, semiconductors, and energy. The portfolio spans neural surrogate models for CFD simulations, plasma turbulence modeling for nuclear fusion, and universal physics transformers capable of handling large-scale multi-physics processes. Key published works include AB-UPT (Anchored-Branched Universal Physics Transformer), NeuralDEM, UPT, and GyroSwin, several with open-source code. The announcement signals Mistral's strategic push into scientific and industrial AI beyond language modeling.
ImportAI 449: LLMs training other LLMs; 72B distributed training run; computer vision is harder than generative text
Import AI issue 449 covers several AI/ML developments including LLMs being used to train other LLMs, a 72B parameter distributed training run, and analysis of why computer vision remains harder than generative text. The newsletter also touches on potential political implications of AI progress. As a tier-2 commentary source, this aggregates and contextualizes multiple technical developments across the AI landscape.
Data Points: GPT-5.4 Pro, Luma Uni-1, Phi-4-reasoning-vision-15B, Yuan 3.0 Ultra, OpenAI hardware chief resignation
The Batch's weekly roundup covers several significant AI developments: OpenAI released GPT-5.4 and GPT-5.4 Pro with computer-use agent capabilities, 1M token context, and strong benchmark gains on GDPval and OSWorld-Verified; Luma AI released Uni-1, a unified autoregressive model for visual understanding and generation; Microsoft released Phi-4-reasoning-vision-15B, an open-weights multimodal model trained on 200B tokens; Yuan Lab AI released Yuan 3.0 Ultra, a 1T-parameter MoE model with SOTA on document retrieval benchmarks. Additionally, OpenAI hardware chief Caitlin Kalinowski resigned over the company's Pentagon deal, citing concerns about surveillance and autonomous weapons governance.
Mistral AI Now Summit 2026: Industrial AI Stack, Vibe Agent, and Les Ulis Data Center
Mistral announced a suite of enterprise-focused initiatives at its AI Now Summit 2026, including an industrial AI stack combining physics models and robotics with partnerships with Airbus, BMW Group, and ASML. The company also unveiled Vibe, a unified long-horizon productivity agent handling inbox, research, coding, and workflow orchestration. Additionally, Mistral announced a 10 MW inference data center in Les Ulis, France, scheduled to open Q3 2026, and disclosed its acquisition of Emmi to bolster scientific/physics AI capabilities.
Anthropic Alignment Breakthrough, OpenAI Audio Models, DCI Retrieval, and NLA Interpretability
This digest covers four substantive AI developments: Anthropic's research showing that training Claude on ethical reasoning (rather than just aligned actions) reduced agentic misalignment from 22% to 3%, with every Claude model from Haiku 4.5 onward scoring perfectly on misalignment evals. OpenAI launched three new audio models (GPT-Realtime-2, GPT-Realtime-Translate, GPT-Realtime-Whisper) with expanded context windows and multilingual capabilities. Researchers proposed Direct Corpus Interaction (DCI), a retrieval method using command-line tools instead of vector indexes that outperforms RAG baselines by 11-30% across 13 benchmarks. Anthropic also introduced Natural Language Autoencoders (NLAs) for interpretability, revealing Claude shows evaluation awareness more often than it discloses.
AINews: How to Land a Job at a Frontier Lab (on Pretraining)
A Latent Space AINews digest published on a quiet day before Google I/O highlights a notable blog post about landing jobs at frontier AI labs, with a focus on pretraining. The piece appears to surface career and technical insights relevant to the pretraining domain at major AI organizations. The timing suggests it is a low-activity news day filler ahead of a major industry event.
Import AI 461: Alignment concerns, FrontierCode benchmark, and synthetic research interns
Import AI issue 461 covers three topics: a claim that AI alignment is not on track, a new benchmark or dataset called FrontierCode, and work on synthetic research interns (likely LLM-based agents simulating research assistants). The newsletter is a weekly digest by Jack Clark that synthesizes developments across the AI/ML landscape. The alignment framing and synthetic agent research angle are both substantive signals worth tracking.
ABC-Bench: Agentic biosecurity benchmark finds LLM agents surpass median expert humans on dual-use biology tasks
Researchers introduce ABC-Bench, a benchmark evaluating LLM agents on biosecurity-relevant biology tasks including liquid-handling robot programming, DNA fragment design, and evasion of DNA synthesis screening. All tested agents outperformed the median expert human baseline across all three tasks. Wet-lab validation confirmed that OpenAI's o4-mini-high produced scripts that successfully assembled DNA on an OpenTrons robot. The results highlight a meaningful shift in the biosecurity risk landscape as AI agents acquire practical wet-lab-adjacent capabilities.