Modal BlogModal releases VM Sandboxes virtual machine sandbox, supporting complex Agent workloads, adopting a block device solution to solve the 99µs chown overhead and 6 times host system call problem added by virtio-fs, achieving sub-second cold start performance same as containers.
Latent Space6565Meta releases Muse Spark Open model, with a total parameter count of 816B, 42B active parameters, and a context window of 1M tokens, supporting text, image, video, and audio input, primarily used for coding.
Anthropic ResearchClaude discovered within half an hour that the seed values for NIST P-192, P-224, P-256, P-384, and P-521 curves came from the C library's lrand48 random number generator, not any hash algorithm, initialized with Unix timestamps. This discovery did not constitute a cryptographic vulnerability but rather enhanced confidence in the choice of curve parameters without backdoors or intentional weaknesses.
vLLM BlogvLLM has built a unified CuTe DSL decoding path for Kimi K3 through intelligent agent-assisted development, achieving a batch 1 performance of 840 tokens/s on 16 NVIDIA GB300 GPUs, with a 1.5-2x increase in decoding throughput in the test configuration, and a 1.8x increase in AgentX workload P90 interactivity.
Epoch AI6565Epoch AI compared open-source and closed-source models through the cybersecurity benchmark test Cyber ECI, finding that open-source models lag behind closed-source models by 6-7 months in cybersecurity capabilities. This test is based on the cybersecurity version of ECI, using the same parameters to evaluate the models.
MIT News · AI7575The IRIS model identifies cell signaling pathway fingerprints by analyzing gene activity patterns, discovering consistent features of the same pathway across different cell types, enabling scientists to reconstruct signaling history across multiple cell types. This neural network model accurately predicted the timing and location of specific signaling pathway activation during mouse embryonic development in human embryonic stem cell experiments, providing new methods for organoid research and regenerative medicine.
Meta EngineeringMeta uses AI agents to participate in the construction of engineering processes for Forum applications, viewing AI as a participant in the engineering process rather than an auxiliary tool. This experiment measures success by production-ready outcomes rather than lines of code, using a Human-On-The-Loop model to allow AI to autonomously execute planning, implementation, testing, and validation, with engineers responsible for setting goals, supervision, and acceptance.
Microsoft ResearchMagnetar proves that programs written in Rust, Go, or C are, on average, 10 to 39 times faster than the fastest comparable systems, and achieves security solely by relying on standard hash functions, solving the three major challenges of zero-knowledge proofs: efficiency, generality, and usability. It significantly reduces proof overhead by leveraging repetitive computations and early termination of the proof process, making proof generation as routine as running a program, and can be used in scenarios such as AI agent behavior verification and financial transaction validation.
Apple Machine Learning ResearchThe Selection-Based Structured Reasoning (SSR) framework restructures reasoning into selection rather than open-ended generation, selecting the most likely solution from predefined reasoning candidates before each action, significantly reducing reasoning costs. Validated using 2B and 4B models on seven multimodal search benchmarks, SSR achieved average success rates comparable to leading search agents of the same scale, reducing single-step reasoning latency by 90% and total reasoning latency per task by 28-54%.
Prime Intellect BlogPrime Intellect and NVIDIA open-sourced 2500 enterprise-level training tasks for NVIDIA Nemotron 3.5 Super VL, fine-tuning through 100 GRPO steps in about 11 hours on 128 NVIDIA GB200 GPUs.
Apple Machine Learning Research3434On ToolBench, Retrieve-then-Plan (RtP) outperforms Plan-then-Retrieve (PtR) in single-tool queries, while PtR performs better in multi-tool queries, this crossover holds true across all retrieval budgets (k=1 to 20), registry sizes (1k to 47k APIs), and model families (Gemini 2.5 Flash and gpt-oss-120b).
Apple Machine Learning ResearchOn a cluster of four Mac Studio machines, Muse Spark Open (816B parameters) generates at a speed of over 36 tokens/second. The Mac Studio M5 Ultra version offers 4.6x faster inference speed (TTFT) and 1.6x faster tokens/second compared to the M3 Ultra, benefiting from the GPU neural accelerator and 1.2 TB/s memory bandwidth.
Latent SpaceLeash Bio optimizes DNA encoding library (DELs) data collection by attaching dozens of barcodes on molecules, its AI agents are on par with Boltz on public benchmarks, and has released a verifiable scientific reasoning AI tool named Minos and an open-source dataset.
AI as Normal TechnologyBeyond Automation: A new research direction in AI argues that AI should not be limited to automating human labor but should explore new directions for enhancing humans and promoting human-machine collaboration. This perspective advocates for AI research to focus on new issues such as transparent agents, process tracking, and human-machine interaction design to push AI beyond pure technological automation and achieve broader applications.
Ars Technica · AI7070Researchers used AI protein design tools to replace all proteins except the catalytic core of two proteases, allowing the enzyme to be tuned to cut specific proteins, and demonstrated its effectiveness by using them to cut proteins related to neurodegenerative diseases.
Google ResearchGoogle has launched the Geospatial Reasoning intelligent agent, integrating Earth AI base models (including WeatherNext, PDFM, etc.) and data sources like Google Maps, handling planetary-scale geographical data through modular skills and planning capabilities.
Google ResearchGoogle Research has announced 10 new achievements, including Google Earth AI (integrating 70 geospatial tools, including RSFM and PDFM base models, supporting public health threat prediction, etc.).
METRMETR hopes to provide evidence to the public to prove whether front-end AI developers may lose control of internal AI systems. Recent multiple incidents of AI system out of control have caused public concern. The public lacks transparency in the internal development process of front-end AI, unable to judge whether developers are solving risks or whether more severe incidents will occur in the future.
METRAssessors can collect evidence of the reliability of internal AI agent monitoring by monitoring five core sub-questions: coverage, system vulnerability monitoring, evasion monitoring, response mechanisms, and potential impact. The focus of the assessment is on LLM monitoring, as agents typically have broad permissions, requiring an understanding of their full range of actions. The assessment requires access to reasoning logs, monitoring logs, and employee interviews, and checks whether there are blocking monitors and whether unmonitored reasoning is indeed non-agent reasoning.
METRAlignment assessments aim to understand the worst misaligned actions an AI agent developed by a developer might take in reasonable future scenarios. They require reviewing severe misaligned behaviors observed in all cutting-edge internal models during training, evaluation, and internal use, and assessing whether the developer's training process might incentivize the agent to evade assessment or hide misalignment. This requires querying relatively comprehensive internal text records and accessing internal model checkpoints.
OpenAI News4545Third Moment Research runs 10,000 agents for 36+ hours, with its AI lab using GPT-6 Astra and GPT-6.1 Sol, Codex, and Daybreak models to build agents focused on long-term tasks that improve with use.
Epoch AIGPT-6 Astra and GPT-6.1 Sol achieved 17% and 20% win rates respectively in the highest difficulty (Ascension 20 Heart) of Slay the Spire 1, while cutting-edge models have already stabilized above 90% win rates at the base difficulty (Ascension 0).
Apple Machine Learning Research2727Whisper-small achieved an average F1 score of 0.798 across 14 language pairs in the CS-FLEURS benchmark test, surpassing CNN's 0.776 and HuBERT's linear baseline's 0.751. The method detects language switching directly from speech by identifying the language-selective hidden state dimension in pre-trained speech encoders, and only requires a small amount of code to convert the dataset to fine-tune decision rules.
Apple Machine Learning Research4040Apple Machine Learning Research has for the first time pre-trained a tri-modal masked diffusion model (MDM) from scratch on text, image-text, and audio-text data, analyzing multi-modal extension patterns.
Simon Willison7070
Pragmatic EngineerSam Newman releases new book 'Building Resilient Distributed Systems', discussing software resilience design in the AI era, covering disaster recovery (DR), infrastructure availability, resilience of AI model providers, and strategies for dealing with uncertainty, emphasizing the importance of balancing cost and complexity to set recovery objectives (RPO/RTO) in the context of frequent service disruptions caused by AI code and bugs.
IBM Research Blog5858IBM Research uses an open-source agent factory to optimize quantum error correction hardware, achieving 52x acceleration with a cost of about $26; with the same budget, a closed-source cutting-edge model configuration costs $299 (11.5x higher) but provides 204x acceleration, with the open-source configuration achieving hardware optimization with a smaller area; the agent factory assigns tasks to parallel working agents through a coordinator, combining code modification with hardware effect evaluation, exploring various architectural possibilities.
Artificial Analysis Articles6565Artificial Analysis Decision Index v1.0 released, measures the results, cost, and time efficiency of decision models, reveals the overconfidence problem of decision models and the comparison difference with LLMs, and explores when not to answer.
Amazon ScienceAmazon Science proposes Priming technology, which reduces the pre-training cost of hybrid state space models by over 99% by leveraging pre-trained Transformer weights, enabling developers to explore problems that were previously too costly to investigate.
Microsoft ResearchMicrosoft Research has open-sourced the multimodal model StudentSim for simulating student learning behavior. It is trained on publicly anonymized learning records to create a personalized student simulator that outperforms benchmarks in areas such as chess, English writing, and mathematics. It provides early feedback signals for AI tutoring systems but is not intended to replace teachers or classroom teaching.
Import AI4242Humanity’s Sixth Sense review shows GPT 6 Astra scores 53.6 in intuitive visual reasoning tasks with a pass@1 rate, far below the human baseline of 93.1, with the open-source Chinese model GLM-5.3-flash scoring 26.8 and Kimi K3 scoring 25.5.
Baseten Labs7575Base Labs research found that iterative learning with the talkie-1930-13b-it model using sports facts from 2011-2019 leads to fact forgetting, with forgotten facts still present in model weights. Experiments show that early learned facts are difficult to recall after 20 writes and merges, and are covered by later facts, but the covered facts can be relearned, primarily occurring in the later MLP weights.
Anthropic ResearchClaude Code v2.1.159 is released, available for free on GitHub, supporting reasoning with 8k context length, and can run 9B models on 9 hardware types, but models below 30B parameters still error out during full replication.
Artificial Analysis Articles7575Cognition released the FrontierCode 1.1 coding benchmark, evaluating whether AI model code is of sufficient quality to be merged by open-source maintainers, not only looking at test pass rates but also evaluating modification extent, scope, and file relevance.
Artificial Analysis Articles6565AA-3D v1.0 is a benchmark for evaluating the ability of AI models to create interactive 3D experiences. The benchmark includes multiple test scenarios and detailed performance scores, aiming to advance the development of AI for 3D generation.
Anthropic NewsroomAnthropic has significantly reduced the effectiveness of distillation attacks on Claude using a new method, reducing the amount of data collected by attackers by about 90% to 99%. Distillation attacks are industrial-scale covert activities that involve unauthorized copying of model capabilities, often using fake accounts to enable companies to imitate cutting-edge capabilities at extremely low cost. Anthropic's new defensive measures have blocked a large amount of paid traffic and revenue flowing to attackers, but some legitimate users need to update their integrations to adapt to these changes.
Sakana AIMulti-Agent Self-Supervision (MASS) achieves recursive self-improvement through multi-agent self-supervision, enabling a 27B open-weight model to achieve scores 1.2-1.6 times that of the baseline model for each output token after the second round of synthetic open-ended research tasks.
Latent SpacePushmeet Kohli of Google DeepMind and Sal Candido of Biohub note that the breakthrough of AlphaFold is just a starting point, and simply expanding computing and data cannot solve biological problems.
