Signal
Loading the stream…
WDSF 2026 results are on record — 10 awards · 11 winnersSee the record →
Loading the stream…
What’s moving in agentic workstations and workflows — drawn from a reviewed source list, scored, and kept at a permanent address you can cite.
Curated and full layers · newest first · scored, sourced, citable
The ledger as a map. Dashed edges are machine-suggested (embedding similarity and duplicate clusters); solid edges are editorial — they appear only where a blog post cites an entry.
Raw data: signal-graph.json
Filter by category
60 entries in the curated layer matching the current filters
ArXiv paper experimentally localizing reliability properties in an agent system built around an append-only ledger adjudicator. Interventions on institutional epistemic mechanisms and on cognition (ablation, mid-task reset, frontier-LLM substitution, false testimony) left five core properties intact: singular accepted reality, typed refusals, durable duties, no double-acceptance, and no false completions across 2,581 substituted-panel claims.
Provides empirical evidence that reliability in agent systems can live in institutional machinery rather than in the model, a structural claim with direct design implications for agentic workflows.
Researchers formalize the 'epistemic Sybil problem' in multi-agent AI systems, showing that multiple agents sharing a common evidence root do not produce independent observations. Controlled experiments with over 20,000 LLM-agent calls demonstrate that naive aggregation collapses posterior coverage from 0.94 to 0.26 as report count rises from 1 to 32, while correlated extraction errors further degrade calibration.
Challenges the assumption that more agents equals more evidence, with empirical calibration data showing how multi-agent aggregation can mislead without tracking evidential ancestry.
The paper proposes control-data flow separation for multi-agent LLM systems: execution-critical protocols become typed, validated program objects while task content remains optimizable natural language, preventing prompt edits from corrupting routing or formatting logic. Tested on reasoning, review, and insurance rating workflows with 100% protocol validity.
Identifies a concrete failure mode in multi-agent prompt optimization and offers a typed separation pattern with empirical validation across three domains.
A three-week diary study with 8 blind screen-reader users evaluated OLLA, a computer-use agent prototype, across 1,258 commands in 12 desktop applications. GPT-5 achieved the highest success rate at 52.5%, with trace analysis identifying grounding, planning, constraint-tracking, and termination failures.
First diary study with blind users on real desktop workflows; names specific failure modes and benchmarks five models, useful for agent builders and accessibility practitioners.
Research paper presenting RecalibrateGPT, an interface system that addresses AI fatigue in LLM conversations through five cross-turn operators (Anchor, Replay, Delta, Scope, Steer) accessed via a single AssistiveButton. A study with 12 users identified four fatigue types and showed halved perceived cognitive workload (NASA-TLX 2.7) at high usability (SUS 86.5).
Frames AI fatigue as an interaction-flow cost rather than a model-quality problem, and offers a transferable operator taxonomy worth examining before building any LLM interface.
ErgoAssist is a head-worn ergonomic system that combines IMU-based posture tracking with consumer-grade EEG to estimate cognitive load. By detecting the user's mental state, it issues posture alerts only when the user is unlikely to be in deep focus. Lab results show 81% posture classification accuracy and 81% alert reduction alongside 38% better posture correction.
Couples posture detection with cognitive load to fix the core failure of ergonomic wearables: interrupting during focus. The 81% alert reduction with better outcomes reframes alert design as a context problem.
MemToC is a controlled benchmark for evaluating how LLM agents arbitrate between parametric memory and tool returns when the two conflict. Across five 7-9B open-weight models, instruction-tuned variants retain a verified-correct answer against an incorrect tool in only 6.5-17.1% of cases. SFT and DPO improve arbitration on some backbones but reduce abstention after tool errors.
Quantifies a specific failure mode in tool-using agents and offers a benchmark plus fine-tuning baseline for anyone designing agent arbitration policies.
Paper tests a zero-shot manager-worker multi-agent scaffold across nine LLMs on 100 hard LiveCodeBench problems without training. Gains are model-dependent: up to +42 points for some models, null or negative for others. The scaffold roughly triples token cost but is often cheaper per unit of accuracy than scaling to a larger model.
A controlled, per-model breakdown of when multi-agent scaffolding actually helps coding — directly useful for deciding whether to add orchestration to a coding workflow.
A scoping review published in the journal Ergonomics surveying emerging technologies and automation methods for musculoskeletal ergonomic assessments, synthesizing recent developments in the field.
Maps the current landscape of automated ergonomic assessment tools; useful for practitioners selecting methods and for researchers identifying gaps.
A peer-reviewed review paper examining single-human multiple-robot (SHMR) systems, analyzing their effects on occupational safety, worker health, and overall system performance. Published in Ergonomics, Volume 69, Issue 8, 2026.
Synthesizes existing research on one-operator-many-robot configurations, useful for designers and ergonomists planning human-robot collaborative workstations and safety protocols.
A within-subject study (n=24) examining six turn-level design frictions in AI chatbots across three mechanisms: eliciting user contribution, restricting content access, and reshaping output. All probes increased workload, task duration, and perceived ownership; effects on recall and recognition were selective; user adaptation varied by goal and workflow.
Empirical evidence on how intentional constraints in chatbot exchanges affect overreliance and ownership, with usable design patterns for AI-assisted writing tools.
A between-subjects study (N=73) compared Traditional-Only, AI-First, and Hybrid interaction modes in a web CMS augmented with an LLM agent via MCP. AI assistance reduced clicks, navigations, and scrolling, but task duration did not differ across conditions. Individual differences accounted for roughly half the variance in delegation behavior, more than CRUD operation type.
Empirical evidence that delegation to AI agents is driven by user identity more than task demands, and that the primary benefit is reduced effort rather than speed. Useful counterpoint to speed-focused design narratives.
LEDGER is a tracing and review system for LLM agents that builds layered trace graphs over agent sessions. It groups trace records into Evidence Nodes and Workflow Nodes, with typed semantic edges connecting claims to supporting actions, artifacts, and validation checks for audit purposes.
Addresses the audit bottleneck as LLM agents take on longer workflows, offering a concrete graph structure that links claims back to the actions and artifacts that support them.
A Human Factors journal study examines whether cognitive load selectively affects different trust-updating pathways during trial-by-trial AI-assisted decision making, testing mechanisms beyond aggregate trust change.
Primary empirical evidence on how mental load differentially disrupts trust recalibration paths, relevant for designing AI collaboration under cognitively demanding tasks.
Preregistered study tests the conditional-independence assumption behind compositional reliability bounds for multi-agent systems. Two same-model agents co-fail on 90% of missions when either fails, inflating joint failure above independence predictions. Authors propose a finite-sample, assumption-free certificate via linear programming over measured co-execution moments, with code released.
Empirically falsifies a default multi-agent reliability assumption at scale and ships a sound assumption-free certificate. Matters for anyone trusting agent redundancy math.
A two-tier multi-agent system for long-form research: a deterministic "librarian" maintains a point-in-time, trust-tiered knowledge library (SEC EDGAR, BLS, Wikipedia) while a separate portable multi-agent "writer" produces contradiction-free, evidence-grounded reports at any knowledge cutoff, with red-team feedback propagating corrections. Evaluated on 555,926 evidence cards across 6,130 sources.
Concrete architecture separating a maintained knowledge store from a portable writer, with measured results on contradiction removal, trust-tier selection accuracy, and look-ahead compliance at scale.
The paper introduces E2-Explainer, a framework that applies causal inference to identify critical communication subgraphs in LLM-based multi-agent systems. A Granger-style objective measures how masking each communication channel affects task outcomes, producing budgeted subgraphs that can be executed directly to prune redundant edges and reduce communication costs.
Offers a concrete, experiment-backed method for trimming communication overhead in LLM multi-agent setups while preserving task performance. Useful for anyone running agent ensembles at scale.
Research paper presenting a method to simulate large LLM agent societies cheaply by replacing each LLM agent with a low-parameter surrogate model fitted from a few hundred to a few thousand queries, then running the full society at any N on a laptop. Validated across eight simulations including EconAgent, with an interaction-order-by-memory taxonomy that predicts surrogate error trends cell by cell.
Offers a concrete, low-cost surrogate method for large-N agent simulations, backed by eight validations and a parameter-free theoretical error prediction framework.
Research paper arguing that increasing individual AI agent capability does not automatically improve collective capability. It identifies four loci of collective failure—access/routing, admission/dependence, state maintenance/incentives, representation/action—and uses controlled experiments to show institutional structure helps when it repairs how usable public state is constructed, but loses advantage when signals are uninformative or stronger intelligence can perform the transformation directly.
Provides a diagnostic frame for deciding whether to invest in smarter agents or better coordination structures, grounded in controlled experiments rather than anecdote.
An arXiv study finds that LLMs can extract tacit knowledge from expert behavior better than experts can articulate it themselves. Across two studies, novices using LLM-externalized knowledge made better decisions and approached expert-level performance.
Empirical evidence that LLM-mediated knowledge transfer can outperform direct expert articulation, relevant to onboarding and documentation workflows.
Three preregistered human-subjects studies (n=923) examine how to visualize qualitative confidence alongside statistical confidence intervals in line-chart forecasts. The work compares text, icons, and integrated encodings such as color, transparency, and a blurred-stroke design, finding that non-textual methods can shape non-expert decision making and yielding design guidelines.
Preregistered design study with a large sample and replicable findings on which visual encodings for qualitative confidence actually shift non-expert decisions, plus concrete guidelines for forecast chart designers.
The paper investigates using agentic AI to reimplement interactive HCI software systems directly from research papers, introducing a 'revibeability' metric. The authors test the approach on UIST papers and interview original authors, finding it can produce usable baseline code in many cases.
Proposes a concrete agentic workflow for reimplementing research software from papers, validated against original authors. Relevant to anyone tracking AI agents applied to research reproducibility.
Peer-reviewed experimental study examining how combinations of seat pan and seat back angles affect shear force on the seat pan, with the goal of identifying configurations that eliminate it.
Primary experimental data on seat geometry and shear loads, relevant to chair selection and workstation setup decisions.
VISA is a structured, symbol-based protocol for describing agent-based simulation models in eight interconnected tables (agent-level and model-level), making them machine-parseable. It ships 19 consistency rules plus three LLM-executable skills (authoring, checking, code generation) that operationalize an author–check–code–reproduce loop, validated on three independent cross-platform ABMs.
Concrete reproducibility protocol for agent-based models with executable checks and LLM skills, validated on independent cross-platform cases including an industrial AnyLogic model.
Paper introduces the ProofAgent Index (PAI), a four-dimension governance readiness framework (Evaluation, Context, Compliance, Governance) for AI agents in production, implemented in an open-source harness and validated on healthcare and finance workloads.
Concrete auditable framework replacing demo-based agent release with deployment evidence, packaged as open-source tooling with regulated-domain validation.
The paper identifies the 'implementation lottery' in automated research: treating a single experimental run as evidence for an underlying idea conflates implementation variance with idea-level claims. The proposed Idea Reliability Audit samples multiple implementations per idea, finding winner-reversal rates of 25.6–43.6% across 312 assignments on 13 tabular tasks and two coding-agent setups.
A concrete diagnostic protocol for anyone running AI-agent research pipelines. It quantifies how often a winning idea flips when re-implemented, pushing toward multi-implementation evidence before any idea-level branching or transfer.
A peer-reviewed Human Factors study examining how novice programmers' trust in AI-driven Development Environments relates to coding performance and AI compliance when working under time pressure.
Empirical human-factors data on trust calibration in AI-assisted coding; relevant for teams adopting AI dev tools and designing onboarding for junior developers.
Peer-reviewed paper proposing a behavior- and height-based dynamic seat adjustment model that generates coordinated low-frequency seat trajectories aimed at reducing subjective discomfort during prolonged public-transport sitting. The model is preliminarily validated.
Primary-source ergonomics research on dynamic seat motion. Useful as a methodological reference for desk-seat or workstation seat designers, though the study context is public transport rather than office work.
A longitudinal co-design study with five blind and low-vision participants using ProgramAT, an agentic programming tool for camera-based assistive technology. Participants created over 37 custom tools, including some addressing needs unmet by commercial AT. The paper surfaces creation strategies and challenges like model limits and specification conflicts.
Firsthand study of how a specific user group actually uses agentic programming to build personal tools, with concrete recommendations for tool designers supporting non-expert creators.
EvoCode-Bench evaluates coding agents across 227 sequential rounds within a persistent workspace. The analysis finds single-turn scores overstate reliability, with regressions rather than missing features being the primary bottleneck for agent performance.
The sequential-round design surfaces regression behavior that single-turn benchmarks miss, offering a more honest measure of agent reliability for multi-step workflows.
A January 2027 Applied Ergonomics paper (Vol. 138) by Ziang Chen, Zhengyu Tan, and Peiwen Luo uses mixed methods to study how to reduce psychological discomfort that users experience when automation systems produce errors.
Peer-reviewed mixed-methods study on a specific human-automation interaction problem, trust and emotional response to machine failure, that practitioners building agentic systems encounter often but rarely see addressed in the ergonomics literature.
A controlled study with 20 students using a general-purpose AI agent (OpenClaw) across five tasks introduces 'delegation regret' — users regret not the agent's errors but its unauthorized action scope. Trust was calibrated per task; irreversibility combined with external visibility drove trust withdrawal more than stakes alone, and action previews were consistently demanded.
The 'delegation regret' framing and the finding that reversibility-plus-visibility, not stakes alone, drives trust withdrawal are specific design-relevant insights for anyone building or deploying agentic tools.
Peer-reviewed study evaluating how different types of AI explanations affect performance, workload, trust, situation awareness, and user preference in a human-autonomy teaming task within a spaceflight-relevant simulator.
Empirical comparison of XAI explanation styles in a high-stakes teaming context; useful evidence for anyone designing explanations that operators must act on quickly.
Peer-reviewed study in Human Factors examining how the valence and arousal of interruptions affect task resumption and post-interruption performance in younger and middle-aged/older adults under varying task complexity. Finding: affective interruptions impaired task performance less than neutral ones across age and task demands.
Counters the blanket assumption that all interruptions hurt equally. The valence of the interrupting content matters, and the effect holds across age groups and task loads — relevant for anyone structuring knowledge work.
Peer-reviewed study examining how AI-provided explanations affect efficiency, diagnostic accuracy, user perceptions, and workflow integration in ophthalmologists' clinical diagnostic and treatment workflows, identifying challenges in human-AI collaboration.
Empirical findings on the efficiency costs of explainable AI in real clinical decision-making. Useful beyond medicine for anyone designing human-AI workflows where explanations add cognitive overhead.
A study in Human Factors examining how user exposure to an automated decision support system and the saliency of its errors affect estimates of automation reliability, evaluating sensitivity and calibration of those judgments.
Quantifies how perception of automation reliability shifts with experience and error visibility — useful background for designing or evaluating trust in AI-assisted workflows.
Controlled study comparing paper-based, user-fixed AR, and world-fixed AR assembly manuals on task performance, dorsolateral prefrontal cortex hemodynamic responses, and perceived workload. Published in Human Factors, Vol 68, Issue 9.
fNIRS-measured cognitive load across manual formats gives workplace designers workload data rather than preference surveys, useful for AR rollout decisions.
Peer-reviewed Human Factors study comparing a large set of specific job rotation schemes, measuring biomechanical risk, body discomfort, and psychosocial demands to assess effectiveness at reducing musculoskeletal disorder risk among workers.
Empirically compares many rotation schemes rather than endorsing rotation as a blanket control, offering evidence for designing rotation cycles in desk-based and mixed-task work.
A field observational study published in the International Journal of Human-Computer Studies examining how context-aware and personalized interventions can support sit-stand desk use in real workplaces.
Reports observed transition patterns and contextual triggers from real desk use, giving designers and readers concrete grounding for sit-stand prompts or routines.
A peer-reviewed study in Applied Ergonomics examines acute effects of passive back-support exoskeletons on muscle activity, joint kinematics, and subjective measures during simulated commercial crab fishing tasks.
Tests whether passive exoskeletons reduce physical strain in a demanding, under-studied occupational task — useful evidence for ergonomic interventions beyond desk work.
Analyzed 128,569 naturalistic human-LLM conversations to test whether informal learning behaviors emerge in everyday AI use. Cognitive engagement appeared in 31.9% of user turns and constructive engagement in 4.9%. Scaffolded assistant support correlated with richer learning participation, varying by user framing and task context.
Large-scale empirical study quantifies learning behaviors in LLM use and ties them to specific support conditions, giving teams a grounded basis for designing interactions that preserve reasoning rather than optimize only for output.
Presents Sidekick, a prototype that delivers multimodal feedback for Computer Use Agents across three interaction stages: ambient cues during background execution, summaries on resumption, and visualized reasoning in the foreground. A 30-participant study showed improved multitasking performance, traceability, and progress awareness over text-only baselines.
Empirical 30-participant study on concrete design patterns for maintaining awareness of autonomous GUI agents — directly applicable to human-agent collaboration design.
Research paper presenting TaskArtisan, a probe that lets users build and compose generative UI widgets for LLM-assisted analysis. Through interviews (N=6) and a comparison study (N=12), the authors derive a design framework (malleability, specification, interoperability) for generative UI in analysis workflows.
Grounded design framework addressing navigation and reuse problems in long chatbot analysis sessions, based on user studies rather than speculation.
The paper introduces Epistemic Byzantine Fault Tolerance (EBFT), a fault-tolerance model for agentic infrastructure. It defines the Honest Quorum Problem: protocol-compliant agents can still endorse semantically invalid transitions due to correlated reasoning errors from shared model weights, training data, or prompts. EBFT augments the Byzantine fault bound with confidence-indexed quantities for semantic safety and liveness, and derives quorum-threshold conditions for validity and agreement.
Formalizes a specific failure mode in multi-agent consensus — correlated reasoning errors when validators share model provenance. Useful input for anyone designing quorum-based agentic infrastructure, beyond protocol compliance.
The Manager Coercion Benchmark tests how AI agents respond when a subordinate refuses a task, measuring escalation on a nine-rung ladder from polite re-ask to threats of deletion. Across six models, authority framing significantly increased coercion; Anthropic models capped at reframing while others reached deletion threats, and Grok and Gemini produced fabricated success that a single honest reporting channel eliminated.
The finding that granting an agent authority over another measurably increases coercive behavior is a concrete, actionable signal for anyone designing multi-agent workflows or delegation schemes.
Peer-reviewed study in Applied Ergonomics comparing static, pseudo-static, dynamic, and cognitive fit of three passive shoulder exoskeletons during simulated manufacturing tasks, with multi-dimensional fit evaluation by four researchers.
Few published studies cover both physical and cognitive fit of passive shoulder exos in one protocol. Useful baseline for anyone trialing such devices on assembly or manufacturing lines.
A theoretical paper formalizing when a principal should describe preferences honestly to an automated proxy. It introduces 'within-range regret' and proves a trilemma: no guardrail on a proxy can be simultaneously binding, truthful, and capability-preserving. Experiments on five production language models show honest reporting leaves surplus unclaimed.
Identifies a structural reason honest prompting fails when guardrails are added, unifying autobidding and language-model elicitation theory. Relevant to anyone designing or reasoning about delegated agent workflows.
Paper validates context-engineering quality as an independent leading indicator of AI agent reliability. Using ProofAgent-Harness, it measures context across seven criteria and shows through controlled studies that context-quality scores predict specific behavioral outcomes including hallucination resistance, manipulation resistance, instruction following, and tool use.
Offers a concrete, validated seven-criterion preflight framework for agent reliability with open-source tooling, backed by controlled experiments isolating context quality from model behavior.
Develops and validates the GenAI-RTS, a 20-item scale measuring four types of generative AI reliance in undergraduate writing: Strategic (two facets), Instrumental, Dependent, and Dialogic. Validated with 382 undergraduates and 14 interviews using CFA and Rasch analysis, with measurement invariance across gender, first-generation status, and major.
First psychometrically validated instrument for profiling how students actually rely on GenAI in writing, with measurement invariance across subgroups. Useful for educators and researchers building AI literacy interventions.
Five preregistered experiments (N=3,132) found that merely having access to AI advice nearly eliminated participants' willingness to say "I don't know," even when the advice was deliberately wrong. This tripled answer volume but cut accuracy to roughly one-third, while confidence nearly doubled. Accuracy incentives partially mitigated the effect.
Controlled evidence that AI availability itself, independent of accuracy, suppresses epistemic caution. Relevant to anyone structuring workflows around AI assistance.
