Meituan Technology Team said more than 20 of its papers were accepted by EMNLP 2026 and presented 12 main-conference long papers. The selection covers streaming voice agents, reinforcement learning, instruction-following evaluation, agent training, text-to-SQL, memory, tool-use data synthesis, neural interpretability, reward modeling, and long-context architectures. The descriptions and experimental results come from the team’s own publication, and the acceptance claims have not been independently confirmed in the supplied material.
Microsoft has open-sourced Scope, a research-preview evaluation platform that lets AI coding agents such as GitHub Copilot and Claude Code run assigned tasks and checks whether they use SDKs, APIs and related tools correctly according to documentation, iThome reported. The platform records each run's operations and test results, along with the AI model and agent version and tool settings used, so teams can see where agents go wrong. Teams can define completion criteria and dependency chains between checks, and Scope uses a large language model to judge whether conditions were met while keeping operation logs and working-directory snapshots for debugging. The project is MIT-licensed on GitHub, must be self-deployed, and offers a web management interface plus command-line and API access; Microsoft advises repeated testing and human review because agents can behave differently across runs and automated judgments may err.
Arnon Mazza and Elad Levi submitted the SCOPE paper to arXiv on October 22, 2026. SCOPE compiles a natural-language policy into an executable program of neural predicates, combining small models, deterministic extractors, and code that composes their outputs. The program and its training data are optimized jointly using adversarial counterexamples targeting deviations from the policy. In the authors’ evaluation on a benchmark of real-world policy tasks, SCOPE achieved the highest average accuracy across tasks and models compared with prompt optimization, task decomposition, and workflow-generation methods. They also report that programs optimized with the weakest executor transferred to other models without further optimization, produced more stable predictions than monolithic models, and generated verdicts interpretable through their predicate outputs.
An arXiv paper submitted on 15 October 2026, "SCOPE: Semantic Bounded Credit Optimization for Prompt Engineering in Text-to-image Generation" by Yufei Jiao, Jinfa Huang and six co-authors, targets reinforcement-learning prompt rewriting for a frozen text-to-image generator. The authors argue that existing methods apply a single image-derived signal to individual subword tokens, even though objects, counts, attributes and relations form coherent semantic spans. SCOPE builds on Group Relative Policy Optimization: the rewriter jointly produces an enhanced prompt and its semantic partition, pools shifts in policy log probabilities within each span to derive a shared coefficient for its token losses, and keeps a single advantage per completion, without separate span rewards or additional image renders. The authors report average relative gains over vanilla GRPO of 9.48% on GeneVal and 7.62% on T2I-CompBench across Qwen-Image, SD3 and FLUX.1-dev, with ablations attributing the gains to both semantic boundaries and within-span pooling.
What problem it solves
Developers need abstractions that can accommodate increasingly agentic LLMs without limiting their reasoning and tool-use abilities through rigid prompts or opinionated orchestration. They also need a way to turn agent components into multi-tenant, multi-session applications with supporting infrastructure.
Key features
Composable agent building blocks — Compose agents from a…
What problem it solves
Readers need a structured way to understand AI Agent design principles and apply them through engineering practice. The book addresses this by organizing the material into 10 chapters and pairing the text with 109 companion experiments covering topics such as context engineering, memory, tools, evaluation, post-training, continuous evolution, and multi-agent collaboration.
K…
What problem it solves
People need a simple way to access and manage files through a web interface. FileBrowser Quantum addresses this with source configuration, user access controls, sharing, search, previews, editing, and a responsive interface.
Key features
Multiple source configuration — Configure multiple file sources with include and exclude rules for browsing and managing different location…
What problem it solves
Security teams need to discover assets, enumerate subdomains, identify exposed information, and scan for vulnerabilities across multiple targets. They also need a way to distribute scanning work across nodes and extend the tool with additional capabilities.
Key features
Asset and subdomain discovery — Map assets, enumerate subdomains, extract URLs, and identify web fingerpri…
What problem it solves
Large language models can be difficult for individuals to train, deploy, or understand because of their scale and the abstractions used by many frameworks. The project addresses this by providing native PyTorch implementations, compact model structures, open datasets, and staged workflows for pretraining, fine-tuning, preference optimization, tool use, and agentic training.
…
What problem it solves
Applications need a consistent way to send and receive WhatsApp messages, manage connected accounts, and forward activity to other systems. Integrators also need configurable event delivery, device-scoped operations, and interfaces suitable for automation clients.
Key features
REST API for WhatsApp messaging — Expose WhatsApp messaging through an HTTP API with a machine-read…
What problem it solves
ChatGPT conversations need a controlled way to work with real project files, terminals, desktop applications, and long-running tasks. Users also need visibility, intervention, recovery, and session continuity when work is delegated across workers or spans multiple chats.
Key features
Local project and terminal control — ChatGPT can read and edit files, run tests, keep termin…
What problem it solves
Generic security prompts and fragmented report collections do not scale across repeated bug-hunting engagements. The project addresses this with structured methodology, vulnerability-specific skills, enterprise attack matrices, validation gates, reporting guidance, and engagement coordination for authorized external testing.
Key features
Topic-triggered security skills — Ski…
What problem it solves
Using Lark/Feishu business capabilities often requires working across many services and API levels, while automation also needs concise interfaces and predictable output. The project addresses these needs with human- and agent-friendly shortcuts, curated API commands, raw endpoint access, and documented agent skills.
Key features
Agent-ready skills — Structured skills cover …
What problem it solves
Agent memory is often opaque: stored context cannot be easily inspected or understood. OpenViking addresses this by organizing knowledge, memory, and skills in a browsable virtual filesystem, with summaries that help agents identify relevant context before reading full files.
Key features
Unified context filesystem — Knowledge, memory, and skills share one filesystem address…
What problem it solves
Teams often spread project context and decisions across chat, source-control tools, CI dashboards, release tools, search systems, and automation. Buzz places these activities in one event log so messages, workflows, code changes, reviews, and approvals share an identity model and audit trail.
Key features
Shared rooms for humans and agents — People and AI agents participate …
What problem it solves
Long-horizon agents need a shared, durable world model that can be queried and updated continuously without becoming tightly coupled to the storage or execution layer. Omnigraph addresses this with typed graph data, multimodal retrieval, reviewable branches, declarative deployment configuration, and server-side policy enforcement.
Key features
Typed ontology — Define node an…
What problem it solves
Image creation often requires switching between separate tools, manually translating visual ideas into parameters, and losing the context of an ongoing conversation. This project brings generation, editing, reference workflows, model comparison, and asset management into a DeepSeek Harness conversation and canvas.
Key features
Chat-Based Image Creation and Editing — Describe…
What problem it solves
Developers often need to install, select, and switch among multiple Node.js versions while working across projects. The project addresses this with a per-user, per-shell version manager that supports common POSIX-compliant shells and command-line workflows.
Key features
Install and switch Node.js versions — Install a selected Node.js release and activate it from the command …
What problem it solves
General-purpose agents can leave changesets partially reviewed, report issues at inaccurate code locations, and produce inconsistent review quality. Open Code Review addresses these limitations by combining deterministic review controls with an agent that can retrieve context and make dynamic review decisions.
Key features
Deterministic review controls — Engineering logic se…
What problem it solves
Data analysis workflows often require expressive query composition, flexible execution models, and interoperability with columnar data. Polars addresses these needs by allowing the DataFrame API and SQL to share the same optimizer and engine, with support for large workloads and custom extensions.
Key features
Unified DataFrame and SQL Queries — Queries can be written with e…
What problem it solves
Codebases need systematic audits and implementation plans, but using an expensive model for routine execution can be inefficient. The skill separates codebase understanding and planning from implementation so execution can be delegated while keeping the plan as the primary output.
Key features
Repository-aware audits — The skill maps the repository's stack, conventions, buil…
What problem it solves
GitHub workflows often require moving between a terminal and GitHub to manage pull requests, issues, and other GitHub concepts. This project places those capabilities alongside git and code, reducing the need for separate browser-based interactions.
Key features
GitHub workflows in the terminal — Manage pull requests, issues, and other GitHub concepts directly from the comma…
What problem it solves
AI coding agents lose context between sessions, so the same decisions get re-explained and nothing about the project is remembered. They also lack a shared workflow and evidence, leaving the human to read every line to confirm what was done. This project adds memory, a workflow, and evidence to agents you already have.
Key features
Persistent project memory with Engram — Eng…
What problem it solves
Coding agents may silently make assumptions, fail to surface tradeoffs, overcomplicate implementations, and modify code they do not fully understand. The guidelines seek more focused changes, earlier clarification, simpler code, and verifiable execution.
Key features
Think Before Coding — The guidelines require explicit assumptions, clarification of ambiguity, consideration …