ALD/E Neuro-symbolic Query Benchmark: 33 Scientific Queries over Machine-Actionable ORKG Comparisons

ALD/E Machine-actionable Table Query Results Validation This survey aims to assess whether machine-actionable versions of review-table data, published in the Open Research Knowledge Graph (ORKG), can provide meaningful and useful insights when queried computationally. Traditional review tables in PDFs are di�cult to search, �lter, or update, which limits their reusability. By converting these tables into a machine-actionable form, we enable queries that can integrate new �ndings, compare conditions across studies, and allow researchers to retrieve insights instantly. Your expert evaluation will help determine which queries genuinely support ALD/ALE research practice and which ones need re�nement. Estimated time: 30 minutes to 1 hour We kindly request that you set aside one uninterrupted hour to complete the survey. Your expertise is essential for assessing whether machine-actionable reviews should become standard practice in the ALD/E community. The respondent's email ([email protected]) was recorded on submission of this form. ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 1 of 45 11/26/2025, 8:49 AM Below is one example of a traditional review table mapped to an ORKG comparison. Each row in the PDF becomes a paper contribution, and each column becomes a structured property that can be queried programmatically. Randall Higuchi less than 1 year between 1 to 5 years between 5 to 10 years more than 10 years Query Evaluation This survey asks you to evaluate a series of natural-language queries executed on machine-actionable versions of review tables modeled as ORKG comparisons. Name * No. of yrs. of educational and working experience in Materials Science * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 2 of 45 11/26/2025, 8:49 AM For each query, please rate: - Meaningfulness: Does the question make scienti�c sense? - Usefulness:Would having this synthesis instantly available (and continually updated with new data) save you time compared to manually consulting the original PDF tables In total, you will evaluate 19 queries for ALD and 14 queries for ALE. The ALD queries span seven tables drawn from three papers, and the ALE queries span eleven tables drawn from six papers. Most tables have two queries, and in one ALD paper and one ALE paper you will also encounter crosstable queries designed to integrate data across multiple tables. What you will do in this survey: 1. Click the provided ORKG comparison link (machine-actionable table) 2. Click the SPARQL query link 3. Run the query by pressing the blue Run button 4. Review the results at the bottom of the screen 5. Provide your ratings Please Note: You do not need to understand SPARQL. Your evaluation should focus only on: (1) the natural-language question, and (2) the resulting synthesized output You may assume that a future system automatically translates natural-language queries into the appropriate graph-query language. ALD Paper 1 — Saturation profile based conformality analysis for ALD in LHAR channels DOI: https://doi.org/10.1039/D0CP03358H ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 3 of 45 11/26/2025, 8:49 AM Table 2 — Machine-actionable comparison Please open the machine-actionable version of Table 2: https://orkg.org/comparisons/R1469158 The following two queries were implemented on this machine-actionable version of Table 2. Q.1 Run the following query https://tinyurl.com/lhr-reactor Natural-language question being evaluated: “Show all combinations of reactor types and LHAR structures reported in the ORKG comparison, and count how many times each combination occurs across the included studies.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.1 is meaningful. * Q.1 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 4 of 45 11/26/2025, 8:49 AM Q.2 Run the following query https://tinyurl.com/pillarhall3-ctma Natural-language question being evaluated: “At 300 °C in PillarHall-3, what were the cTMA values reported across studies” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Need more details of the reactor used and conditions beyond just temperature If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) Q.2 is meaningful. * Q.2 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 5 of 45 11/26/2025, 8:49 AM ALD Paper 2 — Atomic layer deposition on particulate materials from 1988 through 2023: A quantitative review of technologies, materials and applications DOI: https://doi.org/10.48550/arXiv.2506.17725 Table 3 — Machine-actionable comparison Please open the machine-actionable version of Table 3: https://orkg.org/comparisons/R1469383  The following two queries were implemented on this machine-actionable version of Table 3. Q.3 Run the following query https://tinyurl.com/Phosphor-SiO2-ALD Natural-language question being evaluated: “Which phosphors were coated with SiO₂ in the ORKG comparison R1469383 that represents Table 3 of the review ‘Atomic layer deposition on particulate materials from 1988 through 2023: A quantitative review of technologies, materials and applications’?” Strongly disagree 1 2 3 4 5 Strongly agree Q.3 is meaningful. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 6 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Q.4 Run the following query https://tinyurl.com/Red-Eu2-ALD-thinlowT Natural-language question being evaluated: “Among Eu²⁺-doped phosphors with red emission in the ORKG comparison R1469383 (Table 3 of ‘Atomic layer deposition on particulate materials from 1988 through 2023: A quantitative review of technologies, materials and applications’), which ALD coatings were deposited at temperatures ≤150 °C with optimal thickness ≤20 nm, and what precursor schemes were used?” Strongly disagree 1 2 3 4 5 Strongly agree Q.3 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) Q.4 is meaningful. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 7 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Table 4 — Machine-actionable comparison Please open the machine-actionable version of Table 4: https://orkg.org/comparisons/R1469594   The following two queries were implemented on this machine-actionable version of Table 4. Q.5 Run the following query https://tinyurl.com/less-than-40 Natural-language question being evaluated: “ Which support materials were coated at ≤ 40 °C in the ORKG comparison R1469594 (Table 4 of ‘Atomic layer deposition on particulate materials from 1988 through 2023: A quantitative review of technologies, materials and applications’), and which precursor pairs were used, along with the reported coating thickness?” Q.4 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 8 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.6 Run the following query https://tinyurl.com/pharma-hard-query Natural-language question being evaluated: “Among low-temperature runs (< 70 °C) that produced thin coatings (< 20 nm) in the ORKG comparison R1469594 (Table 4 of ‘Atomic layer deposition on particulate materials from 1988 through 2023: A quantitative review of technologies, materials and applications’), list the support material, precursor pair, deposition temperature, coating thickness, and—when alumina is implied by TMA-based precursors—classify the Al₂O₃ growth-per-cycle (GPC) as slow (< 0.4 nm), average (0.4–1.0 nm), or fast (> 1.0 nm).” Q.5 is meaningful. * Q.5 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 9 of 45 11/26/2025, 8:49 AM Q.11 Run the following query https://tinyurl.com/HiGLoT-query Natural-language question being evaluated: “ List all rare-earth ALD/MLD hybrid �lms that achieve high growth per cycle (GPC ≥ 5 Å) at low deposition temperature (≤ 250 °C). For each entry, report the material system, the metal precursor family (e.g. R(thd)₃, R(dpdmg)₃), the organic precursor, the GPC, and the deposition temperature, and sort the results by GPC.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.11 is meaningful. * Q.11 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 16 of 45 11/26/2025, 8:49 AM Q.12 Run the following query https://tinyurl.com/aldmld-linker-gpc Natural-language question being evaluated: “Group rare-earth ALD/MLD hybrid �lms by organic linker family (terephthalate, pyridinedicarboxylate, naphthalenedicarboxylate, pyrazine-based, other) and compute the average growth per cycle (GPC) for each family, considering only �lms deposited at temperatures ≤ 250 °C. Report, for each linker family, the average GPC and the number of �lms contributing to this average, and sort the families by descending average GPC.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree only one family showed up in the output? Q.12 is meaningful. * Q.12 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 17 of 45 11/26/2025, 8:49 AM Table 5 — Machine-actionable comparison Please open the machine-actionable version of Table 5: https://orkg.org/comparisons/R1469991 The following two queries were implemented on this machine-actionable version of Table 5. Q.13 Run the following query https://tinyurl.com/mosled-high-eqe Natural-language question being evaluated: “ Among Er³⁺ MOSLEDs in the comparison, which host matrices achieved high external quantum e�ciency (EQE ≥ 10%) at the lowest threshold voltage, and what annealing temperatures and lifetimes (emission lifetime τ and operational device lifetime OLT) were reported? Return, for each qualifying host matrix, the EQE, threshold voltage, annealing temperature, τ, and OLT, and sort the results by increasing threshold voltage and, within the same voltage, by decreasing EQE.” Strongly disagree 1 2 3 4 5 Strongly agree Q.13 is meaningful. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 18 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Q.14 Run the following query https://tinyurl.com/mosled-pareto-score Natural-language question being evaluated: “Compute an e�ciency-per-volt metric, de�ned as external quantum e�ciency divided by threshold voltage (EQE/Vol), for all Er³⁺ MOSLED entries in the comparison. Return, for each host matrix, the EQE, threshold voltage, the derived EQE-per-Volt value, and (optionally) the annealing temperature, emission lifetime (τ), and operational device lifetime (OLT). Rank the host matrices by descending EQE-per-Volt, breaking ties by lower threshold voltage and then higher EQE, to highlight materials that deliver high emission e�ciency at low operating voltage.” Q.13 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 19 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.15 Cross-table: Tables 3 and 5. Run the following query https://tinyurl.com/t3t5-lowT Natural-language question being evaluated: “ Show the ALD recipe alongside device performance for Er³⁺-based MOSLED host matrices. Join MOSLED performance (Table 5) with ALD process entries (Table 3) and report host, best EQE, ALD material, metal precursor, co-reactant, precursor family, GPC, and deposition temperature. Sort by decreasing EQE.” Q.14 is meaningful. * Q.14 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 20 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Co-ractant is not captured properly in the table Q.16 Cross-table: Tables 2 and 5. Run the following query https://tinyurl.com/LumMOSLEDquery Natural-language question being evaluated: “Which luminescent materials listed as doped systems in the ALD dopant overview (Table 2) also have MOSLED performance data?” Q.15 is meaningful. * Q.15 result provides a useful synthesis of the insights in the Tables. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the tables. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 21 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree The table format is too long without - organizing it a little more to be easier to read will help the user instead of a giant list Q.17 Cross-table: Tables 3 and 5. Run the following query https://tinyurl.com/t3t5-complex Natural-language question being evaluated: “For each luminescent MOSLED host material in Table 5, retrieve the ALD process parameters from Table 3 and report alongside EQE.” Q.16 is meaningful. * Q.16 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 22 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Its taking the second precursor as co-reactant, but the overall data is helpful Q.18 Cross-table: Tables 3 and 5. Run the following query https://tinyurl.com/t3t5correlation Natural-language question being evaluated: “ Which rare-earth oxide matrices share synthesis (Table 3) and device performance (Table 5), and how do synthesis/annealing conditions relate to e�ciencies and lifetimes?” Q.17 is meaningful. * Q.17 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 23 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.19 Cross-table: Tables 2, 3, and 5. Run the following query https://tinyurl.com/t2t3t5complex Natural-language question being evaluated: “From Table 2, select Er-doped luminescent oxides. Combine synthesis temperatures (Table 3) and EQEs (Table 5) and compute: e�ciency index = EQE / synthesis temperature × 100. Rank materials by this index.” Q.18 is meaningful. * Q.18 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 24 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree ALE Paper 1 — Atomic Layer Etching at the Tipping Point: An Overview DOI: https://doi.org/10.1149/2.0061506jss Q.19 is meaningful. * Q.19 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 25 of 45 11/26/2025, 8:49 AM Q.6 Run the following query https://tinyurl.com/t3-complex-fang Natural-language question being evaluated: “ Group all thermal ALE processes by mechanism archetype and compute distinct materials and mean EPC per group.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree ALE Paper 4 — Physical and chemical effects in directional atomic layer etching DOI: https://doi.org/10.1088/1361-6463/ab6d94 Q.6 is meaningful. * Q.6 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 32 of 45 11/26/2025, 8:49 AM Table 1 — Machine-actionable comparison Please open the machine-actionable version of Table 1: https://orkg.org/comparisons/R1560825  The following two queries were implemented on this machine-actionable version of Table 1. Q.7 Run the following query https://tinyurl.com/t1-semi-ale Natural-language question being evaluated: “ List all semiconductor ALE processes and return modi�cation, removal, and activation types; sort by activation mode.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.7 is meaningful. * Q.7 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 33 of 45 11/26/2025, 8:49 AM good table could be made better by including etch rates if available Q.8 Run the following query https://tinyurl.com/t1-sang-complex Natural-language question being evaluated: “ Group semiconductor ALE processes by activation type and count materials per modi�cation–removal pair.” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) Q.8 is meaningful. * Q.8 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 34 of 45 11/26/2025, 8:49 AM ALE Paper 5 — Anisotropic/Isotropic Atomic Layer Etching of Metals DOI: https://doi.org/10.5757/ASCT.2020.29.3.041 Table 2 — Machine-actionable comparison Please open the machine-actionable version of Table 2: https://orkg.org/comparisons/R1563131 The following two queries were implemented on this machine-actionable version of Table 2. Q.9 Run the following query https://tinyurl.com/t2-metals-high-epc Natural-language question being evaluated: “ List all metal ALE processes with EPC ≥ 2 Å/cycle; return material, direction, modi�cation/removal chemistry, EPC, process temperature, and cycle time; sort by EPC.” If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 35 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.10 Run the following query https://tinyurl.com/t2-metals-complex Natural-language question being evaluated: “Group metal ALE processes by direction and compute number of metals and mean EPC per group; sort by mean EPC.” Q.9 is meaningful. * Q.9 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 36 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree ALE Paper 6 — Atomic Layer Etching of SiO₂ for Nanoscale Semiconductor Devices: A Review DOI: https://doi.org/10.5757/ASCT.2024.33.1.1 Q.10 is meaningful. * Q.10 result provides a useful synthesis of the insights in the Table. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 37 of 45 11/26/2025, 8:49 AM Tables I and II — Machine-actionable comparisons Please open the machine-actionable version of Table I: https://orkg.org/comparisons/R1560949 Please open the machine-actionable version of Table II: https://orkg.org/comparisons/R1560977 The following querywere implemented on this machine-actionable versions of the tables. Q.11 Cross-table: Tables I and II. Run the following query https://tinyurl.com/pap6-crossq1easy Natural-language question being evaluated: For each �uorocarbon precursor system appearing in both tables, list precursor chemistries, removal gas, process temperature, etching rate, and ion-energy window (RT ±20 °C), sorted by etching rate Strongly disagree 1 2 3 4 5 Strongly agree Q.11 is meaningful. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 38 of 45 11/26/2025, 8:49 AM Strongly disagree 1 2 3 4 5 Strongly agree Tables III and IV — Machine-actionable comparisons Please open the machine-actionable version of Table III: https://orkg.org/comparisons/R1561025 Please open the machine-actionable version of Table IV: https://orkg.org/comparisons/R1561023  The following querywere implemented on this machine-actionable versions of the tables. Q.11 result provides a useful synthesis of the insights in the Tables. If the amount of data in original table feels small, imagine the same query operating over a much larger, continuously updated version of the tables. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 39 of 45 11/26/2025, 8:49 AM Q.12 Cross-table: Tables III and IV. Run the following query https://tinyurl.com/pap6-crossq2easy Natural-language question being evaluated: “ Collect all anisotropic SiO₂ ALE processes based on C₄F₈/Ar plasma across both tables; list target selectivity pair, selectivity range, chamber-wall treatment, and etch rate; sort by etch rate. ” Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Could be related to the papers but the table doesn't have much useful data - there is a lot of overlap/ repeat values Q.12 is meaningful. * Q.12 result provides a useful synthesis of the insights in the Tables. If the amount of data in original tables feels small, imagine the same query operating over a much larger, continuously updated version of the table. * If your usefulness rating was below 3, please briefly state the reason (one sentence or phrase) ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 40 of 45 11/26/2025, 8:49 AM Tables I, II, and III — Machine-actionable comparisons The following querywas implemented on this machine-actionable versions of the tables. They are linked above. Q.13 Cross-table: Tables I, II, and III. Run the following query https://tinyurl.com/pap6crossq1-complex Natural-language question being evaluated: Group anisotropic SiO₂ ALE processes by �uorocarbon precursor family (e.g. C₄F₈, CHF₃, C₃F₇OCH₃ isomers) and compute mean/max etching rate, union of ion-energy windows, and max selectivity. Strongly disagree 1 2 3 4 5 Strongly agree Strongly disagree 1 2 3 4 5 Strongly agree Q.13 is meaningful. * Q.13 result provides a useful synthesis of the insights in the Tables. If the amount of data in original tables feels small, imagine the same query operating over a much larger, continuously updated version of the table. * ALD/E Machine-actionable Table Query Results Validation https://docs.google.com/forms/u/4/d/1RRxu-wqUK8V-CA1bed_reK1... 41 of 45 11/26/2025, 8:49 AM