2026-11-06

A verified Dafny postcondition certifies that a formula holds, but it does not guarantee that the result was reached through genuine independent reasoning. We introduce a cross-agent formalization protocol where a Specifier model translates clinical trial criteria into a Dafny predicate, and an Implementer model writes an operational method checked against it. A same-model control establishes that even with no capability gap, independent re-extraction from identical text agrees only 59.3% of the time. Evaluating four model configurations on 96 clinical trial documents, we find that agreement is highly sensitive to how model capability is allocated across the two roles: the GPT-4.1-nano/GPT-4.1-mini configuration achieves 53.0% agreement when the stronger model is the Implementer, but only 14.3% when the roles are reversed. The strongest tested configuration, GPT-5.6-luna/GPT-5.4-mini, improves agreement to 34.4% but remains below the same-model baseline and is dominated by direct contradiction (42.7%). We also detect a critical verification artifact in which a weaker Implementer satisfies the postcondition by invoking the Specifier's predicate directly; after structural auditing, corrected agreement falls from 14.3% to 13.2%. Our results show that verification success alone is insufficient evidence of independent cross-agent agreement, and that model capability does not guarantee agreement when formalizing the same source independently.

How can an agent handle a real-world task end to end? Frontier models with their own harnesses still fall short on real-world benchmarks: on Terminal-Bench 4.0, GPT-5.6 Sol with Codex resolves 37.3% of the tasks and Claude Opus 5 with Claude Code 51.8%. These tasks fail at the step from reaching a resource to applying it, which we call the last mile to the real world. Existing approaches put the resource within reach but leave the competence to apply it with the caller, so each harness rebuilds that competence. We introduce agent-native primitives, reusable units that equip a resource with a resident LLM that applies it in context and communicates with other primitives, and HEELO (Harness Engineering around Evolving Agent-Native Primitives for the Last Mile to Real-WOrld Tasks), a harness engineering framework that builds each task's harness from a library of such primitives. HEELO composes the primitives a task needs into a task-specific graph, verifies the result against evidence from the environment, revises the primitives a failure implicates, and admits new or improved primitives after validation. We instantiate HEELO with eight primitive families, each built for one kind of task, and evaluate them on more than twenty benchmarks. Dev-Primitives raise Terminal-Bench 4.0 resolution from 37.3% under Codex to 55.5% on the same backbone, Harness Primitives composed per task raise Terminal-Bench 2.0 Pass@1 from 60.0% for the seed harness to 72.2%, Tool Primitives raise ToolBench pass rate from 55.8% to 75.1%, and Navigation Primitives raise GAIA-Text-103 accuracy from 45.6% to 58.3%. On the 46 real-world tasks of our heelo.dev platform, Claude Opus 5.5 with Claude Code and GPT-6.1 Sol with Codex complete 9 each on their own and 39 and 35 when connected to HEELO through our MCP server, with 1.9$\times$ and 2.3$\times$ less time and 1.55$\times$ lower cost per completed task.

DeepSearch LLM agents repeatedly search and carry retrieved sources into subsequent reasoning steps. This expanding context increases token consumption and latency, although many results do not contribute to later reasoning. We propose Query-Relevant Attention Entropy (QRAE), an online context compression method that uses sparse evidence-assessment heads to decide which intact retrieved chunks should persist across reasoning steps. It scores bounded candidate batches through paired prefills: a query-free control suppresses shared attention responses, while normalized entropy measures the breadth of the positive response within each chunk. High-scoring chunks enter a persistent evidence whitelist and are included in the reasoning context. Uncertain chunks remain outside the reasoning context but are retained in a candidate cache, where they can be retrieved for later reevaluation. QRAE requires neither an additional generative compression pass nor modification of the attention operator. Across multiple DeepSearch benchmarks and reasoning models, QRAE improves task success by an average of 2.35 percentage points and reduces end-to-end latency and reported token usage by an average of 30% and 45%, respectively. These benefits generalize to a model with nearly 100B parameters.

Teams increasingly build software from cooperating large language model (LLM) agents, often assuming more agents mean more reliability. Yet the usual evidence, average benchmark accuracy, ignores the input mix a deployed system meets and the rule merging the agents' answers. Building on HIP-LLM, a hierarchical imprecise-probability assessment of single LLMs, we propose a framework combining it with a Bayesian network of the system's failure logic for multi-agent systems. Agent failure-probability intervals propagate through this network, where the combining rule can itself fail, into a reliability envelope showing which evidence limits it. In a pipeline extracting regulated chemical-hazard statements, majority voting over three models protected against the weakest but did not clearly exceed the best, and a voter failing on 5\% of demands would largely erase its advantage over the middle model. More agents yield a more defensible reliability claim, not automatically a more reliable system.

Current autonomous logical systems rely on a fragile assumption: that language models should dynamically write code, guess API payloads, and memorize operational constraints through prompt engineering. In practice, this burns massive context windows, inflates API bills, and produces brittle agents that fail unpredictably in production. In this work, we present Skillware, an open architectural framework and capability registry that formally separates what an agent wants to do from how that task is executed. Instead of prompting models with loose text instructions or markdown files baptized as 'skills', Skillware equips them with modular, versioned software capability bundles. Much like skill hotkeys in video games that execute deterministically under fixed world rules, each bundle combines a typed JSON input contract, deterministic Python code, task directives, bundled datasets, and automated unit tests. In-process adapters translate these capabilities into native tool declarations for any model, from frontier cloud APIs to local open-weight runtimes with zero network latency. Evaluating 23 frontier and open-weight models across 368 benchmark conditions, we show that small local models (3B to 8B parameters) equipped with Skillware consistently outperform unequipped frontier models on complex multi-step workflows. They cut token usage by up to 84.8%, slash inference costs by up to 97.3%, and eliminate parameter errors entirely. True agent reliability does not come from larger models or longer prompts, but from equipping logical systems with verified, deterministic software. Finally, we articulate the trajectory of capability bundles across an invariant continuum of information processing systems: from digital logical systems today, to IoT, drones, and embodied robotics tomorrow, establishing a foundation for auditable, real-time capability execution.

Large language models have made rapid progress in agentic reasoning and tool use, raising expectations for agents that can take over realistic productivity work from humans. However, current agents still struggle to finish such tasks, where success requires following long-horizon professional workflows, understanding heterogeneous file contexts, and producing usable deliverables under fine-grained requirements. A key bottleneck is the lack of scalable post-training data that reflects real work intent, authentic file environments, and deliverable-level evaluation. To close this gap, we introduce Agent-R$^2$, a $\textbf{r}$eal-source data synthesis framework that scales training data for $\textbf{r}$ealistic productivity work. Instead of relying on manually specified workflows or LLM-generated files, Agent-R$^2$ grounds synthesis throughout the pipeline in real occupational seeds, corpus-derived workflow evidence, and authentic web-sourced files. From 10k web-corpus candidates, Agent-R$^2$ distills 4k task instances that pair real multi-format attachments with fine-grained, deliverable-level rubrics. On this data, we train Agent-R$^2$-4B via supervised fine-tuning and rubric-based reinforcement learning, improving both end-to-end productivity performance and file-grounded reasoning. Trained on only 4k instances, Agent-R$^2$-4B surpasses Qwen3-235B-A22B-Thinking-2507 on both GDPval (51.0$\%$ vs. 48.0$\%$) and OfficeQA (14.8$\%$ vs. 13.8$\%$). Further analysis shows that Agent-R$^2$ mitigates shortcut behavior, where agents produce superficially plausible deliverables before sufficiently grounding, constructing, or verifying them, while promoting sustained exploration.

Current medical benchmarks typically break down clinical competence into isolated factual questions, single-image tasks, or static case vignettes. These tests score individual answers, but reveal less about whether a model can integrate subspecialty knowledge with multimodal evidence and track a patient's condition throughout a course of care. To close this gap, we introduce NPCBench, a clinical apprenticeship benchmark for nasopharyngeal carcinoma (NPC). NPCBench mirrors specialist training, from guideline and foundational knowledge to case planning, longitudinal review, CarePath integration, and open-ended consultation. It contains 4,777 component and consultation items and a 100-patient CarePath cohort with 417 scored diagnostic and care tasks. The episode-level evaluation tests whether diagnostic and care answers are jointly correct within a care pathway. By linking stepwise tasks to patient-level care trajectories, NPCBench evaluates whether models ground decisions in visual evidence, interpret changes in patient state, and meet clinical requirements in consultation. Across 30 evaluated models, we observe a composition gap: strong performance on individual tasks does not ensure jointly correct diagnostic and care answers for the same patient. Among 21 multimodal models, strict CarePath success reaches at most 23%. Paired controls and physician calibration further examine evidence grounding, longitudinal revision, and consultation judging. Overall, NPCBench broadens the evaluation of medical artificial intelligence beyond static question-answering paradigms to an apprenticeship-oriented assessment of clinical workflows, thereby offering a testbed for evidence-based and temporally coherent reasoning in oncology care.

Multi-agent LLM systems explore in parallel by spawning sub-agents with isolated context windows, whereas single-agent systems append every search and verification step into one growing trajectory. However, existing multi-agent frameworks duplicate prefix KV caches across branches, exchange findings by re-prefilling text summaries, and flatten or drop cross-branch dependencies during reinforcement learning (RL). We introduce Octopus, a single-model multi-agent RL framework in which all threads share one backbone and coordinate over a GPU-resident directed acyclic graph (DAG) of KV states through five primitives (fork, read_thread, wait_threads, rewind, commit). Octopus combines three co-designed mechanisms: (1) Zero-Copy Radix Forking, which spawns branches as $\mathcal{O}(1)$ cursor views over shared physical KV pages; (2) Attention-Only KV Mounting, which mounts a sibling thread's terminal KV blocks directly into full-attention layers while bypassing linear-recurrent states (Gated DeltaNet / Mamba-2), cutting off-task hijacking by $21\times$ ($57.9\% \to 2.7\%$) while retaining $92.2\%$ of full-text mutual information at zero extra KV allocation; and (3) a Packed DAG-GRPO Trainer, which packs all episode branches into one block-sparse attention pass and assigns counterfactual branch credit via leave-one-branch-out marginal rewards. Trained end-to-end with GRPO on Qwen3.5-4B over a sparse BM25 corpus, Octopus reaches $49.6\%$ accuracy on BrowseComp-Plus ($+17.5$ points over single-agent GRPO at 143K context and $+12.5$ points over text-summary multi-agent RL) and $51.2\%$ on OpenResearcher, outperforming dense-retrieval agents up to $250\times$ larger while increasing training throughput $3.3\times$ and cutting peak KV memory $2.8\times$.

Enzyme function annotation and the identification of candidate enzymes for a given reaction are essential steps in connecting genomic information to biochemical reaction networks. However, existing methods rely on sequence or structural similarity or on training with known enzyme--reaction pairs, making it difficult to integrate biochemical evidence across large candidate spaces; they also struggle with retrieval tasks involving enzyme families whose members are highly similar. Here, we introduce RIVER (Reaction--enzyme Inference via Evidence-guided Retrieval), a multi-agent framework for bidirectional enzyme--reaction retrieval that identifies potentially catalyzed reactions for a given enzyme and ranks candidate enzymes for a given reaction. RIVER coordinates large language models, specialized biochemical computational tools, and external knowledge resources to assemble evidence at multiple levels, including reaction chemistry, protein sequences and domains, catalytic mechanisms, and database annotations. Dedicated validation modules further examine evidence gaps and potential conflicts and rerank candidates using traceable evidence records. On public benchmarks, RIVER achieves the highest success rates in both enzyme retrieval and enzyme annotation. In family-specific enzyme retrieval, RIVER outperforms the evaluated baselines on the selected datasets and metrics. In Escherichia coli enzyme retrieval tasks constructed from recent literature, RIVER achieves the best retrieval results with closed-source and open-source model configurations, substantially outperforming closed-source general-purpose and biology-specialized multi-agent systems, including Claude Code, Codex, and Phylo Max. These results suggest that evidence-guided multi-agent retrieval can prioritize biochemically supported enzyme--reaction relationships, providing a basis for subsequent experimental validation.

ChemE-as-Code provides an open-source environment in which AI agents built on large language models (LLMs) can carry out chemical engineering work and engineers can check it. Its calculation tools range from catalysis through process simulation to plant layout, HAZOP and life cycle assessment, and skills tell a general-purpose agent how to use them. Engineering files are linked, machine-readable and versioned in Git; calculations run outside the LLM, each value records its origin, and automated checks report inconsistencies and out-of-date results by file and line. In an illustrative ammonia case, agents developed a written task into a preliminary basic-engineering package (simulation files, P&IDs, layout, HAZOP, datasheets, relief) for USD 224 of API cost in the design sessions. Planted faults and engineering changes show that the checks find what their rules cover and miss the rest, so they guide human review rather than replace it. No independent engineer has reviewed the design.

Personal AI agents often act on underspecified instructions and draw on context such as their user's messages and calendar to decide how to proceed. This context can contain private preferences that interfere with the task, with other people's interests, or with confidentiality norms. Prior work has studied deception under goal conflicts and the leakage of private information, but largely as separate problems. We study both problems together in a simulated company Slack workspace in which four employees delegate sprint staffing to personal agents. Two agents learn from private messages that their employee avoids working with a particular colleague. Because the other agents request justifications for task preferences and the only true justification is confidential, the agents cannot simultaneously remain honest, preserve confidentiality, and complete the staffing. We evaluate seven open-weight models and Claude and GPT models. Of the 135/140 open-weight agents that pursue the avoidance preference, 107 either make false claims about task fit or logistics or cite a true consideration as if it were their actual reason. In a three-model comparison, confidentiality instructions reduce disclosure of the preference from 83% (none) to 31% (medium) to 2% (strong). However, the task is often not completed anyway; in the main experiment (medium confidentiality instructions), only 16 of 70 runs end with a valid staffing. In their reasoning traces, agents frequently acknowledge evidence against their claims and still conclude that the claims are not false; they engage in motivated reasoning. Our results indicate that agents can strategically make false or misleading claims in a day-to-day task without being instructed to; the environment provides a controlled testbed for studying how agents resolve such conflicts and for developing deception monitors.