Evidence (4781 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
9875 claims
Filter claims →
Productivity
8807 claims
Filter claims →
Governance
7870 claims
Filter claims →
Human-AI Collaboration
7560 claims
Filter claims →
Org Design
4892 claims
Filter claims →
Innovation
4781 claims
Filtered →
Labor Markets
4004 claims
Filter claims →
Skills & Training
3308 claims
Filter claims →
Inequality
2332 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 | 870 | 233 | 116 | 1066 | 2363 |
| Governance & Regulation | 976 | 451 | 218 | 133 | 1809 |
| Organizational Efficiency | 949 | 224 | 144 | 88 | 1416 |
| Technology Adoption Rate | 764 | 287 | 141 | 122 | 1325 |
| Research Productivity | 501 | 152 | 74 | 362 | 1101 |
| Output Quality | 542 | 216 | 69 | 69 | 896 |
| Decision Quality | 387 | 198 | 94 | 54 | 740 |
| Firm Productivity | 513 | 67 | 101 | 27 | 714 |
| AI Safety & Ethics | 249 | 303 | 73 | 36 | 667 |
| Market Structure | 190 | 192 | 134 | 27 | 548 |
| Task Allocation | 243 | 77 | 91 | 36 | 452 |
| Innovation Output | 291 | 33 | 55 | 20 | 401 |
| Skill Acquisition | 206 | 72 | 65 | 21 | 364 |
| Employment Level | 133 | 63 | 115 | 22 | 335 |
| Fiscal & Macroeconomic | 153 | 79 | 52 | 32 | 323 |
| Task Completion Time | 206 | 37 | 12 | 15 | 272 |
| Firm Revenue | 179 | 52 | 29 | 5 | 266 |
| Consumer Welfare | 130 | 76 | 47 | 13 | 266 |
| Inequality Measures | 48 | 137 | 51 | 6 | 242 |
| Worker Satisfaction | 101 | 81 | 25 | 13 | 220 |
| Error Rate | 84 | 110 | 11 | 5 | 210 |
| Wages & Compensation | 98 | 47 | 30 | 10 | 185 |
| Regulatory Compliance | 88 | 73 | 17 | 7 | 185 |
| Automation Exposure | 66 | 64 | 33 | 16 | 182 |
| Team Performance | 105 | 29 | 30 | 11 | 176 |
| Training Effectiveness | 109 | 22 | 14 | 21 | 168 |
| Developer Productivity | 114 | 21 | 14 | 8 | 158 |
| Job Displacement | 12 | 90 | 24 | 1 | 127 |
| Hiring & Recruitment | 57 | 9 | 9 | 5 | 80 |
| Skill Obsolescence | 6 | 56 | 9 | 1 | 72 |
| Social Protection | 43 | 17 | 8 | 2 | 70 |
| Creative Output | 35 | 21 | 9 | 4 | 70 |
| Labor Share of Income | 18 | 21 | 17 | 1 | 57 |
| Worker Turnover | 15 | 16 | — | 4 | 35 |
| Industry | — | — | — | 1 | 1 |
Innovation
Remove filter
VR generates high‑value behavioral and biometric datasets for AI personalization, training, and analytics; firms that extract this data can gain competitive advantages, creating incentives to centralize collection unless counteracted by policy or market forces.
Economic implications inferred by the authors from the literature synthesis and standard industrial‑organization logic; not supported by original empirical market data in the paper.
There is a need for regulatory standards, industry best practices, and ethics‑by‑design approaches; interoperable policy frameworks are recommended to govern VR security and privacy.
Policy and governance recommendations synthesized from multiple reviewed studies and the authors' integration; presented as prescriptive guidance rather than empirically tested interventions.
An effective defense mix for VR combines technical controls (secure boot, attestation, encrypted communications), AI tools for anomaly detection and policy enforcement, and human‑centered design (transparency, consent, usable controls).
Cross‑study synthesis showing these categories recur as recommended controls in the 31 reviewed papers; authors propose combining them in TVR‑Sec. No deployment or performance metrics provided.
Socio‑Behavioral Safety measures (moderation, design constraints, psycho‑social safeguards) are necessary to prevent harassment, persuasion, addictive interfaces, and other psychological harms in shared virtual spaces.
Qualitative synthesis of social‑behavioral harms and proposed mitigations reported across the literature review (31 studies); comparative evaluation of socio‑technical controls.
User Privacy in VR requires managing highly sensitive behavioral and biometric traces with privacy‑preserving ML approaches (e.g., federated learning, differential privacy), consent mechanisms, and data minimization.
Repeated recommendations across the reviewed studies; authors synthesized privacy‑preserving technical approaches and governance mechanisms from the 31‑study corpus. No primary experiments demonstrating efficacy provided.
System Integrity defenses should cover hardware, firmware, sensors, and networks to protect against spoofing, device tampering, malware, and supply‑chain attacks.
Aggregated technical recommendations from the literature corpus (31 studies) and the authors' mapping of integrity threats to controls (secure boot, attestation, encrypted communications). No empirical testing of these controls in the paper.
The Three‑Layer VR Security Framework (TVR‑Sec) integrates System Integrity, User Privacy, and Socio‑Behavioral Safety into an adaptive, multidimensional defense architecture for VR systems.
Conceptual synthesis developed by the authors from a comparative literature review of 31 peer‑reviewed studies (2023–2025); framework created by mapping identified vulnerabilities to technical, AI, and human‑centered controls. No empirical validation or deployment testing reported.
A coordinated Omnibus that clarifies interactions with the DSA and establishes consistent AI-focused enforcement capacity can reduce regulatory frictions, lower compliance costs, and better align incentives for responsible AI deployment.
Policy recommendation based on comparative mapping and scenario analysis; qualitative argumentation rather than empirical testing.
The iterative, human-in-the-loop agent workflow enables evaluation and refinement of algorithmic clusters into logically consistent, theory-ready categories.
Described iterative loop where agents evaluate clusters, align semantics, and refine outputs; qualitative assessments reported though no formal user-study metrics included in summary.
DAFT via LoRA reshapes semantic vector geometry to highlight domain-relevant distinctions without full model retraining.
Methodological claim: LoRA fine-tuning applied to foundation embeddings to adjust vector space; no geometric analyses or quantitative illustrations provided in the summary.
Across six domains THETA outperforms LDA, ETM, and CTM on measures of coherence and domain interpretability.
Reported comparative experiments across six domains using coherence metrics and qualitative/human interpretability assessments against LDA, ETM, CTM. Summary does not provide effect sizes, statistical tests, or per-domain breakdowns.
THETA substantially improves the interpretability and domain-specific coherence of topic/cluster outputs on very large social-text corpora.
Reported experiments comparing THETA to traditional topic models (LDA, ETM, CTM) across six domains; evaluation reportedly used topic coherence metrics and human-in-the-loop interpretability assessments/qualitative comparisons (no numeric results provided in summary).
Lowering fixed costs via shared resources can enable more entrants and niche innovators (e.g., specialized clinical apps).
Workshop economic implications and participant assertions in breakout sessions and plenary at the NSF workshop (Sept 26–27, 2024).
Public investment in shared data and compute as nonrival public goods will reduce duplication, lower entry barriers, and increase total R&D productivity.
Workshop implications for AI economics articulated by participants and authors as a policy recommendation; rationale stated in the summary document (NSF workshop, Sept 26–27, 2024).
De-risk pathways from lab to clinic via reproducible benchmarks, continuous monitoring, and cross-sector collaborations (academia, industry, clinicians, regulators).
Workshop translation-focused recommendations and roadmap produced by consensus at the NSF workshop (Sept 26–27, 2024).
Enable safe, accountable, and resilient platforms (including virtual–physical healthcare ecosystems) to reduce translational risk.
Workshop recommendations addressing safety, resilience, and virtual–physical ecosystems from cross-disciplinary discussion at NSF workshop (Sept 26–27, 2024).
Promote scalable validation ecosystems grounded in objective, continuous measures and physics-informed models.
Workshop validation and safety theme recommendations from panels and consensus-building exercises (NSF workshop, Sept 26–27, 2024).
Develop clinic workflow–aware systems and human–AI collaboration frameworks to fit real clinical practice and decision chains.
Stated systems and workflows recommendation from expert panels and clinician participants at the NSF workshop (Sept 26–27, 2024).
Build shared compute infrastructures tailored to medical workloads and validation needs.
Workshop recommendation from infrastructure-themed sessions and consensus outcomes (NSF workshop, Sept 26–27, 2024).
Sustain investment in shared, standardized data infrastructures (datasets, ontologies, benchmarks) to support medical algorithm–hardware co-design.
Workshop infrastructure call presented during breakout sessions and final recommendations at the NSF workshop (Sept 26–27, 2024).
Principal recommendation: shift from isolated algorithm or hardware efforts to integrated algorithm–hardware–workflow co-design for medical contexts.
Stated workshop recommendation derived from panels and cross-disciplinary consensus at the NSF workshop (Sept 26–27, 2024).
Sustained public investment and new validation, governance, and translation ecosystems are needed to de-risk commercialization and accelerate safe, accountable clinical adoption.
Workshop principal recommendation based on qualitative synthesis of expert judgment from participants and breakout outcomes (NSF workshop, Sept 26–27, 2024).
Enabling next-generation medical technologies requires a fundamental reorientation toward algorithm–hardware co-design that is clinic-aware, validated continuously, and backed by shared data and compute infrastructures.
Consensus recommendation from a two-day NSF workshop (Sept 26–27, 2024) in Pittsburgh convening interdisciplinary participants (academic researchers in algorithms and hardware, clinicians, industry leaders). Methods: expert panels, thematic breakout sessions, cross-disciplinary discussions, consensus-building. Documentation at https://sites.google.com/view/nsfworkshop.
Automation of routine SE tasks suggests measurable productivity gains at team and firm levels, but quantification requires causal, outcome-based studies (e.g., throughput, defect rates, time-to-market).
Interpretation of literature review findings and survey-reported perceived productivity gains; no causal empirical estimates provided in the paper.
Empirical survey evidence shows generally positive perceptions of AI tools among software engineering professionals and growing adoption.
Cross-sectional survey of software engineering professionals asking about current tool usage and perceived benefits (productivity, quality, speed); absolute respondent count and sampling frame not provided in the summary.
ML enables predictive features in software engineering: effort estimation, defect prediction, work prioritization, and risk forecasting that support Agile planning and continuous delivery.
Literature review of ML-for-SE research and practitioner survey reporting use or expectations of predictive features; specific model performance metrics or dataset sizes not reported in the summary.
NLP techniques improve requirements management and team collaboration by extracting intent from natural-language artifacts (tickets, specs, PRs) and reducing miscommunication.
Synthesis of prior studies in the literature review and survey responses indicating perceived improvement in requirements handling and communication; survey sample size not reported.
The method lowers the technical barrier for adopting surrogates in economics by removing dependence on specialized Bayesian neural-network techniques while preserving rigorous uncertainty quantification.
Argument in Implications section: decoupling uncertainty quantification from network architecture allows use of deterministic NNs with MCMC-sampled parameter inputs; no user-study or adoption metrics provided.
The theoretical diagnostic (linking distribution mismatch to performance loss) gives practitioners a practical tool to detect when a surrogate trained on one parameter distribution will underperform after recalibration or policy changes.
Paper-provided theoretical result and suggested diagnostic use; empirical validation of the diagnostic is implied but not detailed in the summary.
This approach dramatically reduces computation (training and/or evaluation wall-clock time) compared to approaches that sample network weights (Bayesian NNs) or exhaustively explore parameter grids.
Computational evaluation reported in the paper includes empirical examples demonstrating substantial reductions in wall-clock training/evaluation time relative to weight-sampling or exhaustive-parameter-grid baselines (exact datasets, runtimes, and sample sizes not detailed in the summary).
Training a deterministic neural surrogate conditioned on MCMC-drawn parameter samples reproduces the original (forward) model's uncertainty quantification while avoiding embedding parametric uncertainty inside the network weights.
Methodological description: surrogate is a deterministic NN whose inputs include parameter vectors drawn by MCMC from the model-parameter posterior; uncertainty is recovered by repeatedly evaluating the trained surrogate on those MCMC draws. Empirical examples are reported (details not provided here) showing reproduction of model uncertainty.
The proposed pipeline (CFD -> CFM -> CFR) forms a closed loop that can assess and improve color fidelity in T2I systems.
Paper describes end-to-end workflow: CFD provides training/validation labels for CFM; CFM produces scores and attention maps for evaluation and localization; CFR consumes CFM attention during generation to refine images. The repository contains code implementing the pipeline.
Color Fidelity Refinement (CFR) is a training-free inference-time procedure that uses CFM attention maps to adaptively modulate spatial-temporal guidance scales during generation, thereby improving color authenticity of realistic-style T2I outputs without retraining the base model.
Method description in paper: CFR uses CFM's learned attention to identify low-fidelity regions and adapt guidance strength across space and denoising steps (spatial-temporal guidance). The authors evaluate CFR on existing T2I models and report improved perceived color authenticity; no retraining of base T2I models is required (implementation and code available in the repository).
CFM aligns better with objective color realism judgments than existing preference-trained metrics and human ratings that favor vividness.
Empirical comparisons reported in the paper: CFM scoring shows improved alignment with CFD-based color-realism labels and with evaluation criteria that prioritize photographic fidelity, outperforming preference-trained metrics and the biased patterns in human ratings (paper reports both qualitative and quantitative gains; specific numerical improvements and test set sizes are provided in the paper/repo).
The Color Fidelity Metric (CFM) is a multimodal encoder–based metric trained on CFD to predict human-consistent judgments of color fidelity and to produce spatial attention maps that localize color-fidelity errors.
Model architecture and training procedure described: a multimodal encoder trained using CFD's ordered realism labels to output scalar fidelity scores and spatial attention maps indicating where color fidelity issues occur. Training supervision comes from CFD's ordered labels (paper includes training/validation procedures; exact training dataset splits are in the paper/repo).
Varying sample size, injecting contaminated data, and including algorithm-reconstruction tasks during training allow networks to automatically inherit those properties (e.g., multi-n behavior, robustness, algorithmic outputs).
Empirical: training regimes described include varying dataset size n, contaminated simulations, and algorithm-reconstruction tasks; experiments reportedly show networks trained with these variations exhibit corresponding behaviors at test time. Specific experimental details (ranges of n, contamination levels) are not included in the summary.
Collapsing (aggregation) layers mimic reduction to sufficient statistics and enforce the desirable structure for set-valued (permutation-invariant) inputs.
Theoretical/design claim supported by architectural description and motivation: collapsing layers aggregate across observations to produce summaries, enforcing permutation invariance; supported indirectly by empirical success in simulations. This is primarily an architectural/representational argument rather than a purely empirical result.
The network can learn to approximate the outputs of iterative estimation algorithms (demonstrated by learning an EM algorithm for a genetic-data estimation task).
Empirical: a genetic-data example where the network was trained (including an algorithm-reconstruction task) to approximate the EM algorithm outputs; evaluation shows qualitative/quantitative match to the iterative algorithm. Evidence is from reported experiments comparing network outputs to EM outputs (e.g., MSE between them).
Training the network with contaminated simulations yields estimators that are robust to contaminated observations at test time.
Empirical: experiments included injecting contaminated data into training simulations; evaluation measured robustness at test time under contamination and showed improved performance relative to networks not trained on contamination. Supported by reported robustness comparisons (metrics like MSE under contamination). Specific contamination rates and sample sizes are not provided in the summary.
A branched neural architecture with collapsing (aggregation) layers that reduce a dataset into permutation-invariant summaries can produce parameter estimates that are exactly finite-sample (i.e., reproduce estimator outputs at finite sample sizes).
Empirical & theoretical motivation: architecture includes collapsing/aggregation layers to implement permutation-invariance and summary reduction; simulation experiments reportedly show the network reproduces reference estimator outputs at finite sample sizes (finite-sample matching). The exact experimental settings (sample sizes, number of replications) are not specified in the summary; evidence comes from simulated benchmarks and comparisons to reference estimators.
A single “summary network” trained in a simulation-only framework can solve the inverse problem of parameter estimation for parametric models by mapping simulated datasets to parameters (minimizing MSE).
Empirical: network trained on simulated datasets (each dataset simulated conditional on a known parameter) with a mean-squared-error (MSE) loss between predicted and true parameter; evaluated on synthetic parametric benchmark problems and a genetic-data example. Specific sample sizes and number of simulations are not stated in the provided summary; evidence is based on the reported simulation experiments and benchmark comparisons.
Fewer expensive evaluations translate directly to lower compute hours and therefore lower cloud/on-premise costs for computational materials or chemistry R&D.
Implication discussed in the paper's implications section: economic argument linking reduced expensive evaluations to lower compute cost; not an experimental result but an economic extrapolation based on the reported reduction in evaluations.
Correct application of the described elements (GP with derivatives, inverse-distance kernels, active acquisition, OT sampling, MAP regularization, trust-region control, RFF scaling) reduces the number of expensive underlying-theory (energy/force) evaluations by roughly an order of magnitude while preserving underlying-theory accuracy.
Empirical claim reported in the paper: benchmarks and experiments on representative potential energy surface problems (specific datasets and numerical results are said to be presented in the paper and accompanying code); summary states an approximately one order-of-magnitude reduction in expensive evaluations with preserved accuracy.
Random Fourier features are used to decouple hyperparameter training from prediction, yielding favorable computational scaling for high-dimensional systems.
Paper describes use of random Fourier features to approximate kernels so hyperparameter fitting can be done largely independently of prediction-time complexity; complexity/scaling claims supported by methodological argument and empirical timings in the paper/code.
MAP regularization via a variance barrier plus oscillation detection prevents surrogate-induced pathologies and non-convergent search behavior.
Paper describes MAP priors (variance barrier) and oscillation-detection diagnostics as regularization and robustness measures; authors report these measures prevent instabilities in surrogate-driven searches in their experiments.
Using Optimal Transport (Earth Mover’s Distance) for farthest-point sampling diversifies the training points in configuration space.
Paper introduces EMD-based farthest-point sampling as an extension and reports its use in experiments; implementation described in methods and code.
Inverse-distance kernels better capture atomic interactions in configuration space than generic kernels for these surrogate models.
Paper argues and uses inverse-distance kernel design to reflect physical interatomic distance dependence; benchmark comparisons reported in the paper (details in main text and codebase).
Gaussian process (GP) surrogates that incorporate derivative observations (e.g., forces) improve the fidelity of the surrogate model and provide better local estimates of gradients and Hessians.
Paper describes GP regression with value and derivative observations used to constrain the surrogate; experiments/benchmarks reported in the paper and code demonstrate use of derivative observations in surrogate training (exact datasets and sample sizes referenced in paper/code).
Practical modalities exist for efficient classical estimation of gradients for the covered loss classes: using the classical-approximation machinery to compute analytic gradients or unbiased estimators, finite-difference approaches, and surrogate methods; the paper discusses sample complexity and noise considerations.
Methodological discussion in the paper outlining specific gradient estimation approaches compatible with the classical-approximation results, together with complexity/sample-complexity remarks. This is a methods/algorithmic claim supported by analysis rather than empirical benchmarks.
The paper constructs a single-hyperparameter family of BSBMs that monotonically interpolates from weak expressive power up to full universality, enabling a controlled trade-off between simplicity and expressivity.
Explicit one-parameter family construction and monotonicity argument/proof in the paper showing that increasing the hyperparameter increases expressivity and approaches universality. This is a theoretical construction rather than empirical measurement.