Evidence (8974 claims)
Search and filter individual claims pulled from the papers. Looking for a specific finding ("what's the effect on wages?"), you're in the right place. Want to compare whole outcome categories against each other instead? Use the Evidence Explorer.
The board below groups claims two ways: by broad theme (nine paper-level topics) and by outcome category (the 34 claim-level outcomes that the Explorer and Syntheses also use).
Browse by theme
Nine broad, paper-level topics. Click one to filter the claims below.
Adoption
10085 claims
Filter claims →
Productivity
8974 claims
Filtered →
Governance
8062 claims
Filter claims →
Human-AI Collaboration
7749 claims
Filter claims →
Org Design
5057 claims
Filter claims →
Innovation
4896 claims
Filter claims →
Labor Markets
4088 claims
Filter claims →
Skills & Training
3372 claims
Filter claims →
Inequality
2377 claims
Filter claims →
Claims by outcome category
Counts by direction of finding. These are the same 34 outcome categories the Explorer compares and the Syntheses are written for. A linked row has a published synthesis.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 882 | 244 | 117 | 1097 | 2424 |
| Governance & Regulation | 1010 | 469 | 229 | 135 | 1875 |
| Organizational Efficiency | 977 | 235 | 149 | 90 | 1462 |
| Technology Adoption Rate | 781 | 299 | 143 | 128 | 1362 |
| Research Productivity | 506 | 155 | 74 | 363 | 1110 |
| Output Quality | 555 | 219 | 71 | 70 | 915 |
| Decision Quality | 395 | 200 | 95 | 54 | 751 |
| Firm Productivity | 523 | 67 | 101 | 27 | 724 |
| AI Safety & Ethics | 262 | 309 | 75 | 36 | 688 |
| Market Structure | 195 | 201 | 135 | 30 | 566 |
| Task Allocation | 248 | 77 | 96 | 38 | 464 |
| Innovation Output | 300 | 34 | 55 | 20 | 411 |
| Skill Acquisition | 207 | 75 | 65 | 21 | 368 |
| Employment Level | 138 | 67 | 119 | 24 | 350 |
| Fiscal & Macroeconomic | 156 | 80 | 53 | 33 | 329 |
| Task Completion Time | 211 | 38 | 13 | 16 | 280 |
| Firm Revenue | 183 | 52 | 29 | 5 | 270 |
| Consumer Welfare | 131 | 77 | 48 | 13 | 269 |
| Inequality Measures | 50 | 141 | 54 | 9 | 254 |
| Worker Satisfaction | 104 | 85 | 25 | 13 | 227 |
| Error Rate | 87 | 112 | 11 | 5 | 215 |
| Automation Exposure | 69 | 69 | 37 | 20 | 198 |
| Wages & Compensation | 102 | 49 | 31 | 11 | 193 |
| Team Performance | 115 | 30 | 30 | 11 | 187 |
| Regulatory Compliance | 88 | 74 | 17 | 7 | 186 |
| Training Effectiveness | 109 | 22 | 14 | 21 | 168 |
| Developer Productivity | 116 | 21 | 15 | 8 | 161 |
| Job Displacement | 12 | 92 | 26 | 1 | 131 |
| Hiring & Recruitment | 57 | 12 | 9 | 5 | 83 |
| Skill Obsolescence | 6 | 59 | 10 | 2 | 77 |
| Social Protection | 43 | 17 | 8 | 2 | 70 |
| Creative Output | 35 | 21 | 9 | 4 | 70 |
| Labor Share of Income | 18 | 23 | 17 | 1 | 59 |
| Worker Turnover | 15 | 16 | — | 4 | 35 |
| Industry | — | — | — | 1 | 1 |
Productivity
Remove filter
We propose a novel Co-Regulation Design Agentic Loop (CRDAL), in which a Metacognitive Co-Regulation Agent assists the Design Agent in metacognition to mitigate design fixation.
Methodological contribution presented in the paper (proposed system architecture). No empirical sample size reported for the proposal itself.
We propose a novel Self-Regulation Loop (SRL), in which the Design Agent self-regulates and explicitly monitors its own metacognition.
Methodological contribution presented in the paper (proposed system architecture). No empirical sample size reported for the proposal itself.
Policy efficacy varies significantly across corporate profiles, with the strongest effects observed in non-state-owned enterprises, high-tech firms, and firms located in eastern regions.
Heterogeneity analyses reported in the study (subgroup analysis by ownership, technology intensity, and geographic region).
The estimated positive effect of the pilot zones on corporate NQPF is robust across a comprehensive battery of robustness and endogeneity tests.
Paper reports multiple robustness and endogeneity checks (details not provided in abstract) that reportedly do not overturn main findings.
Mechanism analysis identifies three systemic transmission pathways for the policy: optimizing factor allocation, deepening digital technology empowerment, and promoting green innovation and sustainability.
Mechanism analysis reported in the study (methods not detailed in abstract) attributing the policy effect to three pathways.
The pilot zones create an optimized 'digital environment' that underlies the positive impact on corporate NQPF.
Empirical analysis in the paper attributes improved corporate NQPF to an optimized digital environment created by the policy intervention; mechanism analysis referenced.
The DML approach flexibly controls for high-dimensional confounding variables and functional form misspecification, enabling highly rigorous causal inference compared with traditional linear models.
Methodological claim based on use of Double Machine Learning in the study (described as addressing high-dimensional confounders and misspecification).
Establishment of China’s National Digital Economy Innovation and Development Pilot Zones significantly enhances corporate New Quality Productive Forces (NQPF).
Quasi-natural experiment using Double Machine Learning (DML) framework applied to A-share listed companies over 2015–2023; empirical results reported as statistically significant.
AlphaFold represents an 'oracle' breakthrough in AI for scientific discovery.
Cited as an example of an algorithmic breakthrough that changed a specific scientific subtask (protein structure prediction). The paper frames AlphaFold as a milestone in the history reviewed; no new experimental data presented.
Phase Three employs AI for comprehensive sensitivity analysis while humans provide strategic interpretation.
Descriptive claim about the third phase of the framework and its use in the paper's applied test; presented as the intended role split between AI (computational sensitivity tasks) and humans (interpretation).
Phase One leverages AI for rapid market research aggregation and preliminary pro forma generation.
Descriptive claim about the first phase of the proposed three-phase framework as presented in the paper; conceptual rather than a separate empirical finding.
The framework achieved seventy-one to ninety percent time reduction while maintaining analytical quality comparable to traditional methods.
Empirical result reported from the controlled ChatGPT-4 test on the single 150-unit scenario comparing time to complete underwriting tasks versus traditional methods.
This research develops and empirically validates a three-phase framework for AI-augmented multifamily underwriting through controlled testing with ChatGPT-4 using a standardized 150-unit development scenario in Seattle's Greenwood neighborhood.
Controlled testing described in paper: use of ChatGPT-4 on a single standardized 150-unit development scenario in Seattle Greenwood to evaluate the proposed three-phase framework.
Generative artificial intelligence demonstrates significant promise for efficiency gains across financial services.
Introductory assertion in paper; general statement about the potential of generative AI, not directly derived from the paper's controlled test.
Empirical findings demonstrate that digitalization significantly boosts efficiency and competitiveness of industrial production.
Correlation and regression analyses reported in the study linking digitalization measures to indicators of efficiency and competitiveness across levels of analysis.
Digital technologies (automation, IIoT, ERP systems, AI applications) reduce nonproductive costs, increase per-worker output, and improve the cost-efficiency of production in Kazakhstani enterprises.
Case studies and real examples from named enterprises (Asia Auto, Karaganda Foundry and Engineering Plant, Eurasian Resources Group) presented in the article.
The number of employees and working time have a positive but limited effect on labor productivity.
Results from the study's correlation and regression analysis comparing labor input measures (employee count and working time) with productivity outcomes.
Digitalization is the key driver of labor productivity growth in Kazakhstan.
Empirical correlation and regression analysis reported in the study across enterprise, industry, and national economy levels.
Opportunities arising from cyborg workflows include hyper-personalized narratives, democratized production, and ethical augmentation of underrepresented voices.
Forward-looking/interpretive claim in the paper describing potential benefits and opportunities; conceptual rather than empirically demonstrated in the excerpt.
Scalability is addressed via edge computing to support cyborg workflows.
Design/architectural claim in the paper mentioning edge computing as a scalability mechanism; no deployment-scale measurements reported in the excerpt.
The proposed workflows include robust bias mitigation strategies.
Paper asserts bias mitigation approaches are included and demonstrated in case studies; no quantitative fairness metrics or evaluation details provided in the excerpt.
Cyborg workflows produce enhanced creative output via iterative human–AI refinement.
Qualitative claim supported by case studies and examples presented in the paper (no quantitative creativity metrics or sample sizes reported in the excerpt).
Empirical evaluations validate 25-60% improvements in key metrics.
Paper states empirical evaluation results with a 25–60% improvement range; specific metrics, methods, and sample sizes are not provided in the excerpt.
Case studies in content generation, news curation, and immersive production demonstrate efficiency gains of up to 3x in throughput.
Reported results from unspecified case studies described in the paper; numeric claim provided but case study sample sizes and methodological details are not reported in the excerpt.
The paper proposes a comprehensive framework encompassing modular architectures, hybrid protocols, and real-time collaboration interfaces informed by cognitive science, AI engineering, and media studies.
Architectural and methodological proposal described in the paper (the claim is descriptive of the proposed system; no quantitative evaluation of the framework components provided).
Cyborg workflows fuse human judgment with agentic AI autonomous systems capable of goal-directed planning and execution.
Conceptual description and framework proposed in the paper (no empirical sample or trial details reported).
RL-based AVs improve average fuel efficiency by about 1.86% at lower speeds (below 50 km/h) compared to the IDM.
Macroscopic-level fuel efficiency comparison between RL-based AV model and IDM in simulation, stratified by speed (<50 km/h). Number of simulation runs not stated.
RL-based AVs improve average fuel efficiency by about 28.98% at higher speeds (above 50 km/h) compared to the IDM.
Macroscopic-level fuel efficiency comparison between RL-based AV model and IDM in simulation, stratified by speed (>50 km/h). Number of simulation runs not stated.
Transitioning from fully human-driven to fully RL-controlled traffic can increase road capacity by approximately 7.52%.
Macroscopic simulation experiments producing Fundamental Diagrams comparing fully human-driven traffic to fully RL-controlled traffic. Exact number of simulation scenarios or replicates not provided in the claim text.
This study implements a Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm to control AVs and trains it using the NGSIM highway dataset to enable realistic interaction with human-driven vehicles.
Methodological description in the paper: implementation of TD3 and training on the NGSIM dataset. Dataset referenced but no numeric sample size reported in the claim text.
The result is evidence-based triggers that replace calendar schedules and make governance auditable.
Claimed outcome of applying the decision-theoretic framework in the paper (argumentative; no empirical deployment or case-study evidence reported in the summary).
The paper provides a decision-theoretic framework for retraining policies.
Explicit claim about the paper's contribution; the article presents a framework (conceptual/methodological exposition).
The retraining decision is a cost minimization problem with a threshold that falls out of your loss function.
Decision-theoretic derivation presented in the paper (analytical/theoretical reasoning; no empirical validation reported).
Retraining can be better understood as approximate Bayesian inference under computational constraints.
Theoretical argument and decision-theoretic framing presented in the paper (conceptual/mathematical derivation rather than empirical testing).
The analysis was pre-registered and code and data are publicly available.
Authors' statement in the abstract/paper declaring pre-registration and public release of code and data.
The meta-d' framework reveals which models 'know what they don't know' versus which merely appear well-calibrated due to criterion placement — a distinction with direct implications for model selection, deployment, and human-AI collaboration.
Interpretation and implications drawn from empirical results showing dissociations between calibration metrics and metacognitive measures (meta-d', M-ratio, criterion shifts); argument that this distinction informs practical decisions about model use.
We applied this framework to four LLMs (Llama-3-8B-Instruct, Mistral-7B-Instruct-v0.3, Llama-3-8B-Base, Gemma-2-9B-Instruct) across 224,000 factual QA trials.
Experimental methods reported in the paper listing the four model variants and total trial count (224,000 factual QA trials).
We introduce an evaluation framework based on Type-2 Signal Detection Theory that decomposes these capacities using meta-d' and the metacognitive efficiency ratio M-ratio.
Methodological contribution described in the paper: specification of a Type-2 SDT framework and use of meta-d' and M-ratio as measurement constructs.
The best designs often do not originate from top-ranked ILP candidates, indicating that global optimization exposes improvements missed by sub-kernel search.
Analysis comparing origins of the best final designs vs. their ILP ranking, reported across the benchmark set (12).
Larger gains on harder benchmarks: streamcluster exceeds 20× and kmeans reaches approximately 10×.
Per-benchmark empirical results reported for streamcluster and kmeans in the evaluation.
Scaling from 1 to 10 agents yields a mean 8.27× speedup over baseline.
Empirical evaluation across the reported benchmark set comparing performance with 1 agent versus 10 agents; mean speedup stated in the results.
We evaluate the approach on 12 kernels from HLS-Eval and Rodinia-HLS using Claude Code (Opus 4.5/4.6) with AMD Vitis HLS.
Experimental setup described in the paper reporting evaluation on 12 kernels drawn from HLS-Eval and Rodinia-HLS, using Claude Code (Opus 4.5/4.6) and AMD Vitis HLS.
In Stage 2, the pipeline launches N expert agents over the top ILP solutions, each exploring cross-function optimizations such as pragma recombination, loop fusion, and memory restructuring that are not captured by sub-kernel decomposition.
Method section describing Stage 2 which runs multiple expert agents exploring cross-function optimizations on top ILP solutions.
In Stage 1, the pipeline decomposes a design into sub-kernels, independently optimizes each using pragma and code-level transformations, and formulates an Integer Linear Program (ILP) to assemble globally promising configurations under an area constraint.
Method section describing Stage 1 decomposition, per-sub-kernel optimization and ILP assembly under an area constraint.
We introduce an agent factory, a two-stage pipeline that constructs and coordinates multiple autonomous optimization agents.
Method description in the paper describing the design and implementation of the two-stage 'agent factory' pipeline.
Deployment validation across 43 classrooms demonstrated an 18x efficiency gain in the assessment workflow.
Field deployment described in the paper: system was validated across 43 classrooms and an efficiency gain of 18x in the assessment workflow is reported.
Interaction2Eval achieves up to 88% agreement with human expert judgments.
Reported evaluation results comparing Interaction2Eval outputs to human expert annotations (rubric-based judgments) on the dataset.
Interaction2Eval, an LLM-based framework, addresses domain-specific challenges (child speech recognition, Mandarin homophone disambiguation, rubric-based reasoning).
Methodological description in the paper: a specialized LLM-based pipeline designed to handle listed domain challenges; presented as the approach used to extract structured quality indicators.
TEPE-TCI-370h is the first large-scale dataset of naturalistic teacher-child interactions in Chinese preschools (370 hours, 105 classrooms) with standardized ECQRS-EC and SSTEW annotations.
Authors' dataset construction and description: 370 hours of recorded interactions from 105 classrooms, annotated with ECQRS-EC and SSTEW rubrics as reported in the paper.
All data and models are publicly released.
Statement in abstract asserting public release of datasets and models.