Modern software systems accumulate technical debt over decades of development, which makes migration expensive and largely manual. As coding agents become increasingly capable at bug fixing, can they autonomously perform such migrations? Existing benchmarks cannot answer this question because they evaluate only behavioural correctness, not whether the migration actually occurred. This leads an eas
LLM agents remain unreliable on long-horizon tasks, where small local failures can compound over extended interactions and lead to overall task failure. Although external harnesses can substantially improve robustness, harness design remains a manual and expensive process that requires searching over a large space of prompts, tool configurations, and control logic. We propose AutoSaddler, an autom
Speech-based applications pass spoken queries through automatic speech recognition (ASR) before any retrieval module, so ASR errors enter the pipeline as a fixed upstream constraint. We empirically test whether two extensions to standard retrieval-augmented generation (RAG), entity-graph linking and iterative reformulation, absorb or amplify these errors. Using four English accents synthesized thr
Industrial technical reports contain high-value knowledge for maintenance, troubleshooting, and product engineering, but their heterogeneous structure (dense prose, specifications, tables) makes them difficult to index and reason over with standard retrieval and QA pipelines, and no public instruction-tuning or benchmark datasets are built from such documents. We address this gap with Industrial-I
Concurrent multi-agent coding promises division of labor across modules, robustness through redundancy, and parallel exploration at the natural granularity of multi-file projects. Realtime collaborative editing protocols solve this coordination problem for human teams via Conflict-free Replicated Data Types (CRDTs), but the LLMs underneath generate one token at a time and existing multi-agent codi
As large language models (LLMs) continue to advance in coding capabilities, their potential in cybersecurity has drawn increasing research attention, with closed-source LLMs (e.g., Mythos) delivering advanced cybersecurity capabilities. However, existing open-source efforts remain limited: frontier open-weight models do not provide reproducible cybersecurity training solutions, open-source trainin
Vision-Language-Action (VLA) models can turn multimodal context into robot actions, but their action decoders are still trained largely by behavior cloning. This supervises which motor command was demonstrated while leaving implicit the local objective served by the behavior under the instruction. Future-based supervision enriches action learning with frames, latent observations, trajectories, or
Large Language Models excel at code generation, yet competitive programming exposes a persistent failure mode: existing multi-agent pipelines distribute work over generic planner, coder, and debugger roles and delegate the choice of algorithmic technique to the backbone alone. We present MARS (Multi-Agent Relay of Specialized LLMs), a prompt-only framework in which each agent is a topic specialist
General-purpose language models can reason and synthesize knowledge, but complex work also requires sustained interaction with files, information sources, and executable code, together with state maintenance, failure recovery, and verifiable delivery. We call this working capability: sustained, verifiable progress toward a real-world objective. Apodex 1.1 develops this capability along two complem
We study autonomous mathematical discovery in the Station, an open-world multi-agent environment in which AI agents from different model families pursue a shared research goal without a central coordinator or scripted pipeline. Agents choose their own research directions, conduct experiments, collaborate, and build a shared scientific literature. Across 12 construction problems from the AlphaEvolv
We present a novel approach to efficient LLM harness optimization through adaptive validation task selection. Harness optimization iteratively rewrites the harness code based on validation performance, enabling substantial performance gains without updating the underlying model weights. Existing approaches, however, evaluate a fixed validation set in full at every iteration, incurring substantial
Policy optimization (PO) for Large Language Models faces a stability--exploration trade-off, currently mediated by an action-side Policy-KL regularizer. This puts practitioners in a double bind: keeping Policy-KL constrains response behavior and consumes the action-side exploration budget, while dropping it leaves the optimization without an explicit drift control. We argue for an alternative that
With the rapid progress of diffusion models and large-scale video generation, generative world models are increasingly expected to replace traditional simulators, including physics engines, game engines, and reinforcement-learning environments. Yet the remaining distance from generation to simulation lacks a systematic assessment. We present a capability-based study using an external yardstick: ei
Full-length RNAs, particularly messenger RNAs, often exceed the context lengths used to pretrain existing RNA foundation models, limiting complete-transcript modeling at single-nucleotide resolution. We present RIBOSPAN, a 1.61-billion-parameter bidirectional RNA foundation model natively pretrained with context lengths up to 10,240 nt. RIBOSPAN combines dense bidirectional self-attention, single-
A retrieval-augmented QA system can return different answers after an index expansion even when its requested model identifier, prompt, retrieval policy, evidence depth, rendering, and exposed generation controls are held fixed. Aggregate accuracy may hide these changes when gains and losses cancel, while ordinary generation variability makes one-shot comparisons overstate update effects. We call
Recent work proposes next-chunk reasoning RL for leveraging no-CoT data---corpora such as worked solutions and textbook derivations that contain reasoning-rich content but lack explicit chain-of-thought annotations. The method trains a model to generate implicit reasoning traces and rewards them by their ability to predict the next chunk of text. While promising, existing evaluations primarily com
Search-augmented LLMs increasingly mediate everyday consumer recommendations by retrieving live web content. This creates a new risk: LLM recommenders may consume web content that Generative Engine Optimization (GEO) operators have polluted to mislead them. We ask: to what extent do they become unwitting promoters of fake products? We introduce FORGE (Fake Online Recommendations in Generative Envi
Hint-based reinforcement learning addresses reward sparsity in long-horizon agentic tasks by retaining a prefix of an expert trajectory before each rollout, letting the policy explore from a state closer to success. Its effectiveness hinges on the guidance depth: how much of the trajectory to keep. Existing methods treat this depth as a deterministic scalar. Scheduled approaches share one value ac
Language models are sequential processors, but long-horizon agency requires external information and computation beyond model weights and active context. Prime Agent is an open-source harness for long-horizon evaluation and coding-agent workflows. A persistent IPython REPL follows the Recursive Language Model abstraction for programmatic context processing and test-time compute, while Continual Ha
LLM-based agents execute multi-step tasks, but their behavioral structure remains opaque: long unstructured traces resist the safety auditing and runtime monitoring that deployment requires. Existing approaches operate per-trace or success-only, so they miss the cross-run topology that links next-step and failure prediction. To recover that shared structure, we collapse an entire trace corpus into
774 points, 684 comments
686 points, 289 comments
Interpretability research increasingly asks when concepts emerge during training and whether linear probes recover real structure, but in language models these claims are hard to validate because language offers little ground-truth ordering of concepts or relationships among them. We propose the use of astronomical ground truth through AstroPT, a transformer trained on millions of galaxy images, a
Robot policies receive heterogeneous observations at each decision step, yet sequence models differ in how they organize these inputs over time. We introduce WorldToken, a time-first policy instantiation that fuses multiview images, proprioception, and task conditioning within each policy timestep into one world token. A causal temporal Transformer models the resulting world-token sequence, and a
Agent benchmarks often evaluate only final answers even when agents run on stateful runtimes. We argue this under-specifies what is being evaluated: the proper unit is a declared model-plus-runtime configuration whose failures can occur in evidence acquisition, runtime routing, safety boundaries, or repeated execution. We present ClawProBench, a trace-aware benchmark for runtime-native agent evalu
Autonomous research systems are increasingly capable of executing long research workflows, yet automation alone does not ensure that the resulting process remains scientifically grounded. We introduce AutoResearch, a two-stage system that connects Idea Generation with Idea Execution to address both how research ideas are formed and how they are reliably established through experimentation. In Idea
Machine translation tests masked diffusion language models (dLLMs) because every source token must be rendered faithfully, while fixed canvas decoding must choose target length before denoising. Existing masked diffusion decoding work mainly studies token unmasking order, leaving this length decision under-explored despite its direct effect on coverage and redundancy. We introduce Entropy-Valley (
Indoor immersive communication, empowered by millimeter-wave (mmWave) technologies, is a key enabler for 6G-ready Extended Reality (XR) applications. However, mmWave signals are highly susceptible to dynamic obstacles (e.g., moving humans), leading to frequent line-of-sight (LoS) blockages. Moreover, although massive-antenna systems have been widely adopted indoors, their fixed-site deployment lim
Modern flapping-wing insect-scale air vehicles display agility similar to that of their insect counterparts; however, these impressive maneuvers are only possible with off-board sensors like optical tracking cameras. In this manuscript, we introduce two embedded proprioceptive sensors for insect-scale aerial robots: thin film piezoelectric polymers integrated directly into a driving actuator and a
Turn-taking is a basic organizational feature of human conversation and remains difficult to model in natural, synchronous dialog systems. While existing research has explored multimodal approaches and large language models for turn-ending prediction, there is a lack of naturalistic conversational corpora specifically addressing turn-taking dynamics in Turkish. This study introduces a multimodal T
AI-powered digital avatar streamers must answer product questions, engage viewers, and execute marketing strategies in real time, demanding low latency, frequent strategy updates, and accurate yet effective responses. Evolvable Harnesses, whose Skills, Hooks, prompts, and tools can be updated independently of model weights, enable rapid iteration but expose a trade-off: large models adapt zero-sho
Reliable reward models are essential for text-to-video evaluation and alignment. However, the trade-off between evaluation accuracy and inference efficiency places high demands on the quality of training supervision. Existing approaches often rely on holistic judges with fixed rubrics or open-ended reasoning, leading to incomplete inspection, unfaithful justification, and entangled attribution. We
Correcting health misinformation in dialogue requires more than producing a factual rebuttal: users differ in what they know, what they believe, and what they need to hear, so an effective intervention often depends on first asking the right clarifying question. Yet existing methods either respond immediately or probe indiscriminately, treating clarification as either unnecessary or always benefic
Computer-use agents ground natural-language instructions in screenshots to locate interface elements, yet existing benchmarks do not isolate whether models bind relational language to the correct element. We introduce GUI-Primitives, a 994-item benchmark of contrastive instruction pairs over seven spatial relations in graphical user interfaces (left/right, above/below, containment, alignment, prox
Recent large language models (LLMs) can operate as coding agents that build complete games from natural language requests. Game development is especially demanding because program logic, visual and audio content, interfaces, interaction and playability must function together in one executable artifact. Measuring this capability therefore requires evaluation of both game product and the development
The Agentic Data Operations Platform (ADOP) is a reference architecture on Amazon Bedrock that uses specialized AI agents to automate the full Bronze-to-Silver-to-Gold data pipeline lifecycle, compressing new-source onboarding from weeks to hours while keeping data governance and compliance controls inline.
Give your AI agents governed, auditable access to enterprise tools without consolidating infrastructure. This post walks through a four-scope maturity model (Connect, Control, Catalog, and Harden) for building a governed tool gateway with Amazon Bedrock AgentCore, advancing only when real governance pain demands it.
Generative AI
Input tokens are often a meaningful part of the cost of running Retrieval Augmented Generation (RAG) at scale. This post describes a query-aware context compression pattern on Amazon Bedrock: after retrieval, a smaller model filters retrieved chunks against the query before the primary model answers, reducing input tokens and cost while preserving answer quality.
Panasonic Avionics worked with AWS and the AWS Generative AI Innovation Center to build an agentic AI system on Amazon Bedrock, Amazon SageMaker, and AWS Glue that diagnoses in-flight entertainment and connectivity (IFEC) issues across a global fleet, reducing diagnosis time from hours to minutes while maintaining accuracy.
<div class="block-paragraph_advanced"><p><span style="vertical-align: baseline;">In any organizational behavior class, students will learn that effective delegation is among the most important skills for a seasoned leader. Getting meaningful work done involves careful coordination, starting with a subdivision of projects into manageable tasks, mapped onto the skills of the team, and assigned to th
<div class="block-paragraph"><p data-block-key="eucpw">Welcome to the first Cloud CISO Perspectives for August 2026. Today, Chris Betz explains why the AI era makes it more important than ever to lean into security fundamentals.</p><p data-block-key="3e9dh">As with all Cloud CISO Perspectives, the contents of this newsletter are posted to the <a href="https://cloud.google.com/blog/products/identit
New research introduces three tests to quantify benchmark optimization in speech recognition, finding top open-source ASR models reproduce benchmark transcripts even when the audio contradicts them.
Integrated sensing and communications (ISAC) is a key enabler for uncrewed aerial vehicles (UAVs) in the low-altitude economy. This paper proposes an ISAC waveform that embeds a unique word (UW) into orthogonal chirp division multiplexing (OCDM), termed UW-OCDM, together with corresponding communication reception and cooperative sensing schemes for high-mobility UAV scenarios. For communication, t
Google DeepMind partners with game studios to prototype breakthrough AI gameplay.
322 points, 331 comments
Algorithms & Theory
703 points, 2 comments
Official Cohere technical update and publication covering Energy & Utilities.
Law firms and professional service networks have been using Harvey to build new service models and add value collaboratively.
Official Cohere technical update and publication covering Newsroom.
Harvey’s Legal Intelligence Now Lives Inside Microsoft 365
Centralize Context With Harvey’s Connector Library
Official Cohere technical update and publication covering Open Science.
Official Harvey AI technical update and publication covering Alex Bazin on Empowering Teams with Legal AI.
Official Harvey AI technical update and publication covering How In-House Legal Teams Build the Case for AI Adoption →.
Official Cohere technical update and publication covering Manufacturing.
Official Cohere technical update and publication covering Healthcare & Life Sciences.
Fable 5, Now Available in Harvey
Research Scholars
Harvey's agents now inherit the context of your matters and projects, along with a memory of how you work.
Catalyst Grants
Official Cohere technical update and publication covering Enterprise AI.
GPT-5.6 Sol, Now Live in Harvey
Official Cohere technical update and publication covering AI for Developers.
Customer Stories
A Smarter Inbox Built for Legal Work: The New Harvey for Outlook
Official Cohere technical update and publication covering Financial Services.
Open Science Community
Official Harvey AI technical update and publication covering Theresa Spartichino on Driving Innovation Through Experimentation and Collaboration.
Claude Opus 5, Now Live in Harvey
Analyze the Full Scope of Legal Evidence, Not Just the Documents
Claude Sonnet 5, Now Live in Harvey
Official Cohere technical update and publication covering Product Launch.
All Papers
Official Cohere technical update and publication covering Meet Cohere Transcribe Arabic.
Official Cohere technical update and publication covering Cohere and the University of Toronto partner to advance responsible AI adoption at scale.
Turn Your Standards Into Stronger Reviews
Introducing Harvey Tenet, our first post-trained open-weight model.
Official Cohere technical update and publication covering How CoreWeave used Cohere North to transform its customer support in 90 days.
Official Cohere technical update and publication covering Public Sector.
Official Cohere technical update and publication covering Technology.
Visualize Legal Data on Demand in Harvey
Official Cohere technical update and publication covering Secure AI.
Harvey Launches Ethical Wall Enforcement With Intapp
Official Cohere technical update and publication covering AI Governance.
Future(s) of Work
Official Harvey AI technical update and publication covering Al Hounsell on Driving Legal Transformation That Lasts.
Aya Models
Official Cohere technical update and publication covering Company News.
Press releases and partnership announcements.
Official Cohere technical update and publication covering Hardware-aware dynamic speculative decoding.
550 points, 852 comments
Official technical announcement and publication from Hugging Face covering Measuring benchmark optimization in speech recognition.
We ran 904 DeepSWE rollouts on GLM-5.3 and GPT-5.6 Sol. Sol leads pass@1 by 3.7 points; GLM-5.3 wins pass@4 at half the cost, and a GLM-first cascade hits 85.9%.
Official LMSYS Chatbot Arena release and benchmark update covering Chasing the Batch-1 Floor: Ling-3.0-flash Speculative Decode on Blackwell.
Official technical announcement and publication from Hugging Face covering How Hugging Face Inference Endpoints, Jobs, and Buckets Power Search on Papers with Code.
We ran 904 DeepSWE rollouts on GLM-5.3 and Claude Fable 5. A tie on pass@1, but GLM-5.3 wins pass@4 and costs 5.4x less: \$3.99 per rollout vs. \$21.63.
Official LMSYS Chatbot Arena release and benchmark update covering Fast Engine Recovery: Sub-Second Engine Restart for SGLang via Weight Cache Daemon.
250 points, 196 comments
6542 articles sourced historically · 100 per page