Evidence (4793 claims)
Adoption
5539 claims
Productivity
4793 claims
Governance
4333 claims
Human-AI Collaboration
3326 claims
Labor Markets
2657 claims
Innovation
2510 claims
Org Design
2469 claims
Skills & Training
2017 claims
Inequality
1378 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 402 | 112 | 67 | 480 | 1076 |
| Governance & Regulation | 402 | 192 | 122 | 62 | 790 |
| Research Productivity | 249 | 98 | 34 | 311 | 697 |
| Organizational Efficiency | 395 | 95 | 70 | 40 | 603 |
| Technology Adoption Rate | 321 | 126 | 73 | 39 | 564 |
| Firm Productivity | 306 | 39 | 70 | 12 | 432 |
| Output Quality | 256 | 66 | 25 | 28 | 375 |
| AI Safety & Ethics | 116 | 177 | 44 | 24 | 363 |
| Market Structure | 107 | 128 | 85 | 14 | 339 |
| Decision Quality | 177 | 76 | 38 | 20 | 315 |
| Fiscal & Macroeconomic | 89 | 58 | 33 | 22 | 209 |
| Employment Level | 77 | 34 | 80 | 9 | 202 |
| Skill Acquisition | 92 | 33 | 40 | 9 | 174 |
| Innovation Output | 120 | 12 | 23 | 12 | 168 |
| Firm Revenue | 98 | 34 | 22 | — | 154 |
| Consumer Welfare | 73 | 31 | 37 | 7 | 148 |
| Task Allocation | 84 | 16 | 33 | 7 | 140 |
| Inequality Measures | 25 | 77 | 32 | 5 | 139 |
| Regulatory Compliance | 54 | 63 | 13 | 3 | 133 |
| Error Rate | 44 | 51 | 6 | — | 101 |
| Task Completion Time | 88 | 5 | 4 | 3 | 100 |
| Training Effectiveness | 58 | 12 | 12 | 16 | 99 |
| Worker Satisfaction | 47 | 32 | 11 | 7 | 97 |
| Wages & Compensation | 53 | 15 | 20 | 5 | 93 |
| Team Performance | 47 | 12 | 15 | 7 | 82 |
| Automation Exposure | 24 | 22 | 9 | 6 | 62 |
| Job Displacement | 6 | 38 | 13 | — | 57 |
| Hiring & Recruitment | 41 | 4 | 6 | 3 | 54 |
| Developer Productivity | 34 | 4 | 3 | 1 | 42 |
| Social Protection | 22 | 10 | 6 | 2 | 40 |
| Creative Output | 16 | 7 | 5 | 1 | 29 |
| Labor Share of Income | 12 | 5 | 9 | — | 26 |
| Skill Obsolescence | 3 | 20 | 2 | — | 25 |
| Worker Turnover | 10 | 12 | — | 3 | 25 |
Productivity
Remove filter
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.
CUA-Suite's rich multimodal corpus supports emerging research directions including generalist screen parsing, continuous spatial control, video-based reward modeling, and visual world models.
Authors' claim about potential use-cases and research enabled by the dataset; forward-looking/qualitative statement.
CUA-Suite provides two complementary resources: UI-Vision, a rigorous benchmark for evaluating grounding and planning capabilities in CUAs, and GroundCUA, a large-scale grounding dataset with 56K annotated screenshots and over 3.6 million UI element annotations.
Dataset/benchmark description in paper: UI-Vision benchmark and GroundCUA counts (56,000 screenshots, >3,600,000 UI element annotations).
Continuous video streams preserve the full temporal dynamics of human interaction, forming a superset of information that can be losslessly transformed into the formats required by existing agent frameworks (unlike sparse datasets that capture only final click coordinates).
Argument made in paper contrasting continuous video to sparse screenshots/final click coordinates; conceptual/logical claim about information content and transformability.
VideoCUA provides approximately 10,000 human-demonstrated tasks across 87 diverse applications with continuous 30 fps screen recordings, kinematic cursor traces, and multi-layered reasoning annotations, totaling approximately 55 hours and 6 million frames of expert video.
Dataset description and counts reported in paper: ~10,000 tasks, 87 applications, 30 fps, ~55 hours, ~6,000,000 frames, plus annotation modalities.
Recent work emphasizes that continuous video, not sparse screenshots, is the critical missing ingredient for scaling these agents.
Cites/references recent literature (stated in abstract) asserting the importance of continuous video over sparse screenshots.
Computer-use agents (CUAs) hold great promise for automating complex desktop workflows.
Statement in paper's introduction/abstract; conceptual claim based on prior literature and motivation for the work.
Extensive offline evaluations demonstrate OneSearch-V2's strong query recognition and user profiling capabilities.
Author statement referencing extensive offline evaluations showing these capabilities; no metrics, datasets, or sample sizes provided in the excerpt.
OneSearch-V2 introduces a behavior preference alignment optimization system which mitigates reward hacking arising from the single conversion metric and addresses personal preference via direct user feedback.
Methodological description of an optimization/feedback component in the paper; no empirical quantification of mitigation or user-feedback effects provided in the excerpt.
OneSearch-V2 contains a reasoning-internalized self-distillation training pipeline that uncovers users' potential yet precise e-commerce intentions beyond log-fitting through implicit in-context learning.
Methodological description of the training pipeline in the paper; no direct quantitative evidence or ablation results given in the excerpt.
OneSearch-V2 includes a thought-augmented complex query understanding module that enables deep query understanding and overcomes the shallow semantic matching limitations of direct inference.
Methodological description of the proposed module in the paper; no standalone evaluation numbers for this module provided in the excerpt.
OneSearch-V2 effectively mitigates common search system issues such as information bubbles and long-tail sparsity, without incurring additional inference costs or serving latency.
Author claim in the paper stating mitigation of these issues and no added inference/latency costs; no quantitative measures, benchmarks, or latency numbers provided in the excerpt.
Manual evaluation confirms gains in query-item relevance, with +1.37%.
Reported manual evaluation metric in the paper; no sample size or annotation protocol provided in the excerpt.
Manual evaluation confirms gains in search experience quality, with +1.65% in page good rate.
Reported manual evaluation metric in the paper; no sample size or annotation protocol provided in the excerpt.
OneSearch-V2 increases order volume by +2.11% in online A/B tests.
Reported online A/B test result in the paper; no sample size, test duration, or statistical significance reported in the excerpt.
OneSearch-V2 increases buyer conversion rate by +3.05% in online A/B tests.
Reported online A/B test result in the paper; no sample size, test duration, or statistical significance reported in the excerpt.
OneSearch-V2 increases item CTR by +3.98% in online A/B tests.
Reported online A/B test result in the paper; no sample size, test duration, or statistical significance reported in the excerpt.
OneSearch, as a representative industrial-scale deployed generative search framework, has brought significant commercial and operational benefits.
Author assertion describing OneSearch as industrial-scale and commercially/operationally beneficial; no supporting numerical evidence or sample size reported in the excerpt.
Generative Retrieval (GR) offers advantages over multi-stage cascaded architectures such as end-to-end joint optimization and high computational efficiency.
Statement in paper positioning GR as a promising paradigm and listing these advantages; no quantitative study or sample size reported in the excerpt.
Automation in Japanese manufacturing increased even during periods of slow productivity growth.
Empirical finding from applying the framework to industry-level data in Japanese manufacturing; comparison of inferred automation trends with observed productivity growth periods (exact sample/time not provided in the summary).
Applying the framework to Japanese manufacturing industries shows that automation increased through capital deepening.
Empirical application of the theoretical framework to Japanese manufacturing industries (industry-level analysis); estimation/inference using industry macro observables. (Paper states result; exact sample size/time span not provided in the summary.)
The model provides a transparent mapping from standard macroeconomic observables (capital-labor ratio, output per worker, elasticity of substitution) into the degree of automation, allowing automation to be measured without relying on technology-specific indicators.
Theoretical mapping derived from the CES structure that links observable macro variables to the endogenous degree of automation; methodological claim about inference procedure.
Aggregating task-level decisions generates a CES production function in which the economy-wide degree of automation emerges endogenously.
Analytical derivation in the paper: aggregation of task-level adoption decisions yields a CES aggregate production function with endogenous automation parameter.
The degree of automation is defined as the share of tasks performed by capital rather than labor.
Explicit model definition provided in the paper (conceptual/theoretical definition).
The degree of automation in the aggregate economy emerges endogenously as an equilibrium outcome and can be inferred from standard macroeconomic data.
Theoretical development in a task-based production framework with endogenous technology adoption; mapping from model to observable macro variables (capital-labor ratio, output per worker, elasticity of substitution).
These results demonstrate a practical path toward high-precision, low-latency text-to-SQL applications using domain-specialized, self-hosted language models in large-scale production environments.
Conclusion drawn by the authors based on their implementation, token reduction, and reported accuracy/latency-related claims; generalization to large-scale production is asserted but not supported by detailed production deployment metrics in the excerpt.
The resulting system achieves 98.4% execution success and 92.5% semantic accuracy, substantially outperforming a prompt-engineered baseline using Google's Gemini Flash 2.0 (95.6% execution, 89.4% semantic accuracy).
Reported empirical evaluation comparing the authors' system to a prompt-engineered baseline (Gemini Flash 2.0) with explicit performance percentages for execution success and semantic accuracy; no sample size, test set composition, statistical significance, or evaluation protocol provided in the excerpt.
The approach replaces costly external API calls with efficient local inference.
System design claim: the model is self-hosted and performs local inference instead of using external API-based LLM calls; no cost accounting or latency benchmarks provided in the excerpt.
This reduces input tokens by over 99%, from a 17k-token baseline to fewer than 100.
Reported measurement comparing input token counts before and after applying their approach (explicit numerical baseline and resulting counts provided); no sample size or distribution of token counts reported.
A novel two-phase supervised fine-tuning approach enables the model to internalize the entire database schema, eliminating the need for long-context prompts.
Methodological description (two-phase supervised fine-tuning) and claim that this internalization removes reliance on long-context prompts; no detailed experimental protocol or sample size provided in the excerpt.