News
· 2026-11-16
VisualArtifactSpec 1.0 (VAS) is a benchmark for agents that create or modify visual deliverables. It contains 211 tasks spanning presentations, static graphics, animations and videos, and web interfaces. Each task packages objectives, source materials, constraints, and deliverables, while VAS scores Task Fulfillment separately from Artifact Performance across Content, Organization, Visuals, and Delivery.
Across nine model–harness configurations, overall scores range from 42.12 to 70.52. Artifac
News
· 2026-11-16
VideoNexus is a training-free architecture for long-video agents that moves dense native audio-video inspection into perception workers with isolated contexts. A planner decides where to look, what visual details to resolve, and how densely to sample; workers return compact, timestamp-grounded evidence for iterative verification. With Gemini 3.8 Flash serving both roles, the system reaches 88.6% on full LVBench, outperforming same-backbone static ingestion and built-in agentic processing by 3.6
News
· 2026-11-16
SilentGuard moves semantic validation of tool results out of the LLM and into the harness that executes API calls. The system evaluates each response against API-specific invariants and withholds any result that violates them, preventing silent failures that return HTTP 200 with plausible but incorrect data. Across 15 public research APIs, the study found silent failures in 11, with some counts inflated by as much as 691 times. On targeted faults, wrong-answer rates fell from 74.0% to 1.7% acros
News
· 2026-11-16
Accelerate Bench v1.1 introduces 32 CPU optimization tasks adapted from scientific applications and supercomputing competitions. Rather than optimizing a single kernel, each task gives an agent a working program, workload, correctness contract, timing envelope, and edit boundary; public tests support iteration, while a separate verifier scores the final submission. Across 16 model–harness systems, Score@1 spans 25.5%–67.4%. The two leaders achieve at least 2× speedup on 22 or more tasks; the oth
News
· 2026-11-16
LangChain has announced a $300 million Series C led by CapitalG at a $3 billion valuation. The company framed the round around LangSmith, positioning it as an agent engineering platform that helps enterprises own and operate their intelligence rather than treating agent behavior as an opaque demonstration.
The supplied announcement does not disclose product benchmarks, customer counts, revenue mix, retention, gross margin, or a detailed roadmap for the proceeds. The financing is therefore a sig
News
· 2026-11-16
Amazon Bedrock Managed Knowledge Base and AgentCore Memory can now operate within a single `AgenticRetrieveStream` request. A `memoryConfiguration` binds the memory resource, stable actor identity, and conversation session; the planner can search long-term memory when context is useful, while `userContext` applies document-level ACL filtering to the managed knowledge base. The service also restores short-term history and persists each completed turn, removing much of the application-side convers
News
· 2026-11-16
Financial data-automation provider Duco has rebuilt its agent stack on Amazon Bedrock and Amazon Bedrock AgentCore, combining dedicated per-tenant runtimes and gateways with identity, network, and data isolation. Its most consequential engineering change is an MCP skills architecture that replaces hundreds of statically injected tool schemas with three meta-tools for on-demand discovery and invocation, reducing the main context from roughly 200,000 tokens to 20,000.
Duco reports 10× faster reco
News
· 2026-11-16
AWS presents multi-agent coordination as a distributed-systems design problem rather than a model-scale problem. Its taxonomy separates Tier 1, where code defines the execution path, from Tier 2, where the model selects it at runtime, then maps sequential pipelines, conditional workflows, routers, fan-out/fan-in, plan-based orchestrators, typed handoffs, group chat, and debate/consensus to their operational costs and failure modes.
The manufacturing lens is equally important: p99 latency must f
· 2026-11-16
Axonnic has launched a forearm sleeve that uses 64-channel surface electromyography to decode finger motion and estimate fingertip force while leaving the hand uncovered. During training, EMG is synchronized with hand tracking to create joint-angle labels; the reported inference path relies on EMG alone, with force calibrated through presses on force sensors.
In one reported session, a 215-second recording made four days after training without recalibration produced a 6.9° aggregate error acros
News
· 2026-11-16
LangChain has reorganized LangSmith around a single agent record, positioning one object as the organizing unit for building, running, improving, and governing agents across frameworks and runtimes. The announcement signals a shift from tracing isolated model or tool calls toward managing an agent's longitudinal lifecycle, potentially reducing integration friction between observability, evaluation, feedback, and policy controls.
The supplied announcement does not specify the record schema, supp
News
· 2026-11-16
MoEngage has turned its existing OpenAPI 3.0 APIs into MCP tools with FastMCP and hosts the MCP server on Amazon Bedrock AgentCore Runtime. The resulting Merlin AI stack supports embedded, conversational, custom, and per-customer decisioning agents. Bedrock-hosted foundation models plan tool calls and compose answers, while the MCP server itself remains model-agnostic.
Tenant boundaries are enforced outside the model: AgentCore validates JWTs, server-forwarded headers carry workspace and user c
· 2026-11-16
The Register reports that Google has demonstrated an AI security agent that can infer a fixed defect from a security patch and autonomously produce a working exploit without human assistance. The significant shift is not merely code generation, but the chaining of differential analysis, exploit synthesis, tool use, and execution-based verification into a closed loop.
The supplied material contains only a headline and does not identify the model, target vulnerability, benchmark, completion rate,
News
· 2026-11-16
Microsoft Research, Microsoft Security, and UC Berkeley have released TokenEconomizer, an open-source, black-box plugin that lowers coding-agent token consumption by reusing repository experience across sessions. It neither retrains nor replaces the agent or its underlying model. Instead, it converts prior session logs into one-line micro-skills, prunes tools unlikely to be needed in that repository, and invokes a judge model when an agent appears trapped in a repeated failure loop.
Across 277
News
· 2026-11-16
Novo Nordisk's Manufacturing Technology team has deployed a process-intelligence assistant on AWS for a vault containing roughly four million controlled documents. Amazon Bedrock AgentCore runs the query agent; Amazon OpenSearch Serverless combines BM25 with Titan dense embeddings, while Amazon Neptune stores the process hierarchy, manufacturing entities, document lifecycle state, and cross-document relationships. Amazon Bedrock Rerank then refines the fused candidates before the agent answers w
News
· 2026-11-16
Amazon Bedrock Managed Knowledge Base now adds tabular search to agentic retrieval. Rather than treating a spreadsheet as a set of passage-sized chunks, the service first retrieves the relevant file, lets the agent generate code, and executes that code over the original CSV, XLSX, or XLS file inside an isolated sandbox without network access. Only printed results return to the agent, which combines them with retrieved policy or memo passages and emits a cited answer.
This design restores whole-
News
· 2026-11-16
Amazon Web Services has added managed retrieval instructions to Amazon Bedrock Managed Knowledge Base, allowing each tenant to define a persona, answer format, retrieval guidance, and answer policy while sharing retrieval infrastructure. Instructions are asynchronously reviewed, versioned, and applied through `AgenticRetrieveStream`; Planning trace events expose the exact instruction version used.
The control plane does not replace tenant isolation. Bedrock applies the filter supplied in a requ
News
· 2026-11-16
OpenAI is expanding ChatGPT for Healthcare with computer-use agents, improved clinical answers, and healthcare plugins that can operate across different software systems. The update moves the assistant beyond answer generation toward an execution layer capable of interpreting interfaces, invoking tools, and completing workflow tasks.
The available announcement does not identify the underlying models, supported healthcare applications, rollout scope, benchmark results, or clinical-safety control
· 2026-11-16
WithCoral has released the Apache 2.0 core of Coral, a self-hosted gateway that sits between AI agents and the systems they operate. Agents can turn a successful workflow into a short TypeScript program, save it as a custom tool, and later discover and reuse it alongside standard provider tools. Coral is implemented in Rust and can run as a desktop application or shared server.
Each generated program runs in a V8 isolate without network access or credentials. The gateway checks every tool call
News
· 2026-11-16
This preprint examines a failure mode in LLM-agent memory: a stale-fact resolver can incorrectly hard-delete a fact that the current query needs. Across three models, hard deletion changes from gains of +0.186 to +0.230 under perfect resolution to losses of -0.176 to -0.181 at the lowest tested reliability. Paired interventions show that false drops degrade downstream accuracy significantly more than false keeps, although retaining facts is not risk-free if incorrect labels can steer generation.
News
· 2026-11-15
AURA 2.0 reframes streaming-video understanding as a continuous agent loop rather than a wait-for-prompt, answer-once interface. Model-Initiated Conversation determines when a visually relevant event warrants speaking, Stream-Aware Tool Calling grounds external tool execution in the current video and dialogue context, and Interactive Visual Grounding links responses to user-marked objects through frame-aligned bounding boxes. Long-Horizon Interaction Memory preserves state across extended stream
News
· 2026-11-14
CompoundingKernel reframes GPU-kernel optimization as a verifiable agentic search. Agents receive only a PyTorch definition, while an LLM controller uses a per-problem archive of externally validated kernels and their lineages to choose either a parent or a fresh design for parallel child agents; correctness is decided outside the model. The paper reports site-verified kernels for all 235 tasks and a 4.40x speedup over the benchmark reference on 209 public-board tasks, with its kernels ranked fi
News
· 2026-11-14
A study of agentic on-policy distillation (OPD) shows that the method transfers interaction policies as well as task competence. General-reasoning supervision reduced mathematical tool attempts from 70.4% to 14.4% while raising accuracy from 16.7% to 38.0%. In contrast, ALFWorld supervision drove τ²-bench accuracy from 41.5% to 6.6% and triggered tool calls that violated user-information dependencies.
The proposed selective OPD method ranks source-domain trajectory positions by changes in teach
News
· 2026-11-14
The paper introduces a continuation-guided framework that turns coding agents into offline synthesizers of executable policies. In Craftax, a partially observable and procedurally generated game environment, an agent synthesizes a rule program and refines it through simulator feedback; the resulting program then runs without any inference-time LLM call. The central mechanism is a feedback cascade that replaces a single sparse episode return with hierarchical goals and belief updates, giving loca
· 2026-11-14
SparseRead is a reusable control layer that helps AI agents retrieve only the external evidence needed for a task, without modifying the underlying agent model. Its Read Intent Protocol tracks evolving requirements and source coverage, Adaptive Reader Backends choose how evidence is located, and a Budgeted Read Gate caps cumulative exposure while retaining a faithful full-view fallback.
Across six models and five scenarios, the paper reports up to 92.9% fewer total tokens and 89.0% lower wall t
· 2026-11-14
Anthropic has redesigned Claude Code Projects around parallel AI workers, allowing multiple coding agents to operate on separate parts of a project concurrently and combine their results. This shifts the product from a single-agent coding surface toward a project-level orchestration environment where task decomposition, workspace coordination, context isolation, and result integration must work together.
The update is strategically important because concurrent agents can shorten wall-clock time
News
· 2026-11-13
Continual Recursive Self-Improvement (CRSI) asks an optimizer to carry an evolved harness across compatible benchmarks or policy models. The paper's Self-Improvement Memory (SIM) sits on the optimizer side: it groups edits by recurring harness problem and stores each attempt's rationale, predicted and observed outcomes, and evaluation context. New attempts extend this history as optimization continues.
Across four domains, SIM produced higher final scores than harness-only and root-cause-only c
News
· 2026-11-13
OutfitterBench places LLM agents in a 200-day outdoor-equipment rental business spanning 46 product types and roughly 1,100 reservation requests. Under a fixed capital ceiling, agents must coordinate reusable inventory, outstanding commitments, and uncertain walk-in demand. Across seven frontier LLMs, several agents underperform a simple rule-based baseline; only the leading system consistently exceeds it. Trajectory analysis identifies inconsistent constraint application, decisions that invalid
News
· 2026-11-13
Gamer is presented as a generalist vision-language-action agent rather than a conventional chat model. It combines perception, memory management, and multi-step reasoning, while sharing one training and deployment pipeline across conventional 2D interfaces and complex 3D simulated worlds. Continued pretraining uses tens of thousands of hours of device-level keyboard and mouse trajectories. Sparse-thinking instruction tuning is intended to reduce redundant inference, and a stable multi-turn reinf
News
· 2026-11-13
WebArena-Pro introduces a reproducible multimodal benchmark for web agents: 300 tasks across 20 self-hosted applications in six domains. Text, vision, audio, and video are treated as first-class observations, with dedicated extraction actions for media. Tasks execute in isolation and can be evaluated in parallel; authoring is supported by a dedicated interface, LLM-assisted triage, and final human validation.
The top reported systems are Codex with GPT 6.1 Sol and Claude Code with Claude Opus 5
News
· 2026-11-13
G-CAST converts production traces from an e-commerce agent into a continual self-training loop. It routes examples by reward signal, prioritizes uncertain cases with predictive entropy, and evolves reward definitions under governance constraints so that new demand and harness capabilities can be learned without losing previously accepted behavior.
Candidate checkpoints are promoted only after cumulative evaluation confirms both adaptation to new requirements and regression safety on established