Configuration: HyInv default (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.38 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the composition-filtered reading: sequence entropy at least 3.5 bits, hydrophobic fraction (AVILMFWY) 0.25 to 0.55, longest homopolymer run at most 6 The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source HyInv: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21) Configuration: HyInv default (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.68 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the unfiltered reading, over all designs The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source HyInv: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21) Configuration: ProteinMPNN (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.05 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the composition-filtered reading: sequence entropy at least 3.5 bits, hydrophobic fraction (AVILMFWY) 0.25 to 0.55, longest homopolymer run at most 6 The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source ProteinMPNN: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21) Configuration: ProteinMPNN (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.47 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the unfiltered reading, over all designs The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source ProteinMPNN: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21) Configuration: ProteinZero V2 (ESM2) (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.11 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the composition-filtered reading: sequence entropy at least 3.5 bits, hydrophobic fraction (AVILMFWY) 0.25 to 0.55, longest homopolymer run at most 6 The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source ProteinZero V2: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21) Configuration: ProteinZero V2 (ESM2) (HyInv report) Protocol: Novel and foldable designs per de novo backbone at matched diversity (HyInv report) Dataset: De novo backbones application panel (74 backbones) 0.45 Count (Number of designs per backbone (of 4) that are both foldable (ESMFold2 scTM at least 0.9) and novel (sequence identity below 0.3 to the nearest UniRef50 sequence and to the nearest sequence of the pretraining library), at matched diversity; two readings, with and without the report's composition filter; the unfiltered reading, over all designs The baselines are interpolated at the HyInv diversity from their own temperature grids, so 'at matched diversity' is the report's wording for an interpolated comparison and not a measurement of either baseline at HyInv's diversity.)count · higher
Uncertainty: Not reported by the source
Coverage: 74/74 backbones scored (58 clusters of 30% sequence identity; 4 designs per backbone.)
Author-reported evaluation · Source checked Evidence concern: excluded from comparisons Appendix C states 'We do not claim zero sequence overlap between the training data and the evaluation panels', and Appendix G's homology screen is a coverage rule under which partial hits remain (Table 10: 28 of 74 de novo backbones and 7 of 12 pocket backbones have partial hits, and all 4 named enzymes are homologous to training data by design). Results on these panels are therefore not clean held-out measurements and must not be compared with results from sources that do establish no overlap.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix C 'Data and pool', final bullet; Appendix G 'Homology of the panels' and Table 10
The ten baselines are single published settings at one temperature each, so their design-set diversity differs from HyInv's: on CATH4.3 test HyInv samples at diversity 0.331 against 0.093 for PiFold, 0.104 for CMLM, 0.144 for LM-Design 650M and 0.000 for DPLM-2 3B, which decodes by argmax. Appendix D.2 states that these four differences are not at matched diversity, and Appendix A states that even for the closest baselines the comparisons are 'close to, but not exactly at, equal diversity', the largest offset of 0.016 being worth at most 0.0011 scTM at the development slope. The scTM and scTM difference results of the protocols named here must not be read as a matched-diversity ranking. The results read at a stated diversity level (Tables 8 and 9 and Appendix D.3) are not covered by this entry, because they are the ones the report reads on a fitted curve at a fixed diversity; nor are the recovery, diversity, pLDDT, scRMSD and lDDT results, whose protocols carry the same single-setting caveat in their own limitations and whose metric qualifiers say the value is at each method's own setting.HyInv Technical Report (Hyra Team, 2026-10-14) · Appendix D.1 and D.2; Table 7; Figure 14 diversity block; Section 5.1 'Baselines' and 'HyInv'
The inference-time results were measured with the supervised-stage weights and a rewrite temperature of 0.23, not with the released checkpoint and its default temperatures, and the quality read-outs of the two numerical paths used the RL checkpoint at update 288, not update 640. Speed and quality in this report are therefore measured on different weights, and the 57x figure compares HyInv's optimised BF16 path with ProteinMPNN's released implementation rather than with an equally optimised baseline.HyInv Technical Report (Hyra Team, 2026-10-14) · Table 1 footnote; Appendix F 'Timing protocol' and 'Baseline timings'
Methods, coverage and source ProteinZero V2: novel and foldable designs per de novo backbone at matched diversity (HyInv report)
hyinv-20261106-protocol-novel-and-foldable-yield
Aggregation: Not reported
HyInv Technical Report (Hyra Team, 2026-10-14) · Section 6 'De novo backbones', sentence '0.68 unfiltered and 0.38 filtered per backbone, against 0.47 and 0.05 for ProteinMPNN and 0.45 and 0.11 for ProteinZero V2' (page 8 of 21)