Datasets:
Formats:
csv
Languages:
English
Size:
1K - 10K
Tags:
open-science
superconductivity
superconductor-discovery
strongly-correlated
materials-discovery
2d-materials
License:
Discovery proposals: six C2DB structure hypotheses (Tuli99)
#60
by Tuli99 - opened
- submissions/Tuli99_discovery_1CuS2-1.json +15 -0
- submissions/Tuli99_discovery_1WAs2-1.json +15 -0
- submissions/Tuli99_discovery_1WSb2-1.json +15 -0
- submissions/Tuli99_discovery_2CoTe-4.json +15 -0
- submissions/Tuli99_discovery_2NiPS3-1.json +15 -0
- submissions/Tuli99_discovery_2NiS-2.json +15 -0
- submissions/Tuli99_discovery_2NiTe-2.json +15 -0
- submissions/Tuli99_discovery_2NiTe-3.json +15 -0
- submissions/Tuli99_discovery_2RuS-1.json +15 -0
- submissions/Tuli99_discovery_2RuTe-1.json +15 -0
- submissions/Tuli99_discovery_2SNb2-1.json +15 -0
- submissions/Tuli99_discovery_2SeNb2-1.json +15 -0
submissions/Tuli99_discovery_1CuS2-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "1CuS2-1",
|
| 4 |
+
"osc_id": "OSC-01206",
|
| 5 |
+
"formula": "CuS2",
|
| 6 |
+
"method_name": "Discovery hypothesis: 1CuS2-1; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 22.604,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Polytype twin (p-6m2) of verified OSC-00581 (delta=+0.06).",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_1WAs2-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "1WAs2-1",
|
| 4 |
+
"osc_id": "OSC-01085",
|
| 5 |
+
"formula": "As2W",
|
| 6 |
+
"method_name": "Physics-screened discovery: As2W (1WAs2-1), calibrated N(E_F) ~ 1.16 states/eV/atom from C2DB PDOS. Non-magnetic d-metal, p-3m1 lattice, TM-pnictide chemistry (the material class of the verified top scorers). Conditional dome score 19.5 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 19.455,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p-3m1 lattice, TM-pnictide chemistry (the material class of the verified top scorers); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 1.16; (4) e_hull=0.161 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 20.4; d=0.2222 -> 16.8; d=0.1667 -> 19.5; d=0.1111 -> 16.5. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially.",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|
submissions/Tuli99_discovery_1WSb2-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "1WSb2-1",
|
| 4 |
+
"osc_id": "OSC-01334",
|
| 5 |
+
"formula": "Sb2W",
|
| 6 |
+
"method_name": "Physics-screened discovery: Sb2W (1WSb2-1), calibrated N(E_F) ~ 1.48 states/eV/atom from C2DB PDOS. Non-magnetic d-metal, p-3m1 lattice, TM-pnictide chemistry (the material class of the verified top scorers). Conditional dome score 24.8 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 24.758,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p-3m1 lattice, TM-pnictide chemistry (the material class of the verified top scorers); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 1.48; (4) e_hull=0.352 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 26.0; d=0.2222 -> 21.4; d=0.1667 -> 24.8; d=0.1111 -> 21.0. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially.",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2CoTe-4.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2CoTe-4",
|
| 4 |
+
"osc_id": "OSC-01176",
|
| 5 |
+
"formula": "Co2Te2",
|
| 6 |
+
"method_name": "Discovery hypothesis: 2CoTe-4; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 19.068,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Polytype twin (p-6m2) of verified OSC-00524 (delta=+0.22).",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2NiPS3-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2NiPS3-1",
|
| 4 |
+
"osc_id": "OSC-01318",
|
| 5 |
+
"formula": "Ni2P2S6",
|
| 6 |
+
"method_name": "Discovery hypothesis: 2NiPS3-1; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 23.424,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "MPS3-family analogy (NiPS3 monolayer, near-metal): Ni-S framework as verified OSC-00747; P pair substitution may keep filling near the dome.",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2NiS-2.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2NiS-2",
|
| 4 |
+
"osc_id": "OSC-01321",
|
| 5 |
+
"formula": "Ni2S2",
|
| 6 |
+
"method_name": "Discovery hypothesis: 2NiS-2; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 23.424,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Polytype twin (p4/mmm) of verified OSC-00747 (delta=+0.17).",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2NiTe-2.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2NiTe-2",
|
| 4 |
+
"osc_id": "OSC-01322",
|
| 5 |
+
"formula": "Ni2Te2",
|
| 6 |
+
"method_name": "Discovery hypothesis: 2NiTe-2; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 24.215,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Polytype twin (p4/nmm) of verified OSC-00754: same composition, different polytype; filling and DOS must be recalculated.",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2NiTe-3.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2NiTe-3",
|
| 4 |
+
"osc_id": "OSC-01323",
|
| 5 |
+
"formula": "Ni2Te2",
|
| 6 |
+
"method_name": "Discovery hypothesis: 2NiTe-3; organizer downfolding requested",
|
| 7 |
+
"provisional_score": 24.215,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Polytype twin (p4/mmm) of verified OSC-00754: different polytype; filling and DOS must be recalculated.",
|
| 11 |
+
"at": 1792477047,
|
| 12 |
+
"hypothesis_only": true,
|
| 13 |
+
"estimated_score_source": "Heuristic comparison to a related active material; not a DFT/ED calculation for this proposed structure.",
|
| 14 |
+
"uncertainty": "Composition, polytype or family similarity does not determine active-band filling or DOS. Actual score can be zero. Requires organizer SCF/downfolding with corrected n=2f convention and active-band selection."
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2RuS-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2RuS-1",
|
| 4 |
+
"osc_id": "OSC-00319",
|
| 5 |
+
"formula": "Ru2S2",
|
| 6 |
+
"method_name": "Physics-screened discovery: Ru2S2 (2RuS-1), calibrated N(E_F) ~ 1.17 states/eV/atom from C2DB PDOS. Non-magnetic d-metal, p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate). Conditional dome score 19.6 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 19.564,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 1.17; (4) e_hull=0.05 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 20.5; d=0.2222 -> 16.9; d=0.1667 -> 19.6; d=0.1111 -> 16.6. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially.",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2RuTe-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2RuTe-1",
|
| 4 |
+
"osc_id": "OSC-00812",
|
| 5 |
+
"formula": "Ru2Te2",
|
| 6 |
+
"method_name": "Physics-screened discovery: Ru2Te2 (2RuTe-1), conservative anchor N(E_F) = 1.89 states/eV/atom (verified ceiling; linear DOS extrapolation would give 6.69) from C2DB PDOS. Non-magnetic d-metal, p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate). Conditional dome score 31.6 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 31.626,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 6.69; (4) e_hull=0.161 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). (5) EXTRAPOLATION FLAG: raw DOS per atom 3.48 exceeds the calibration domain max 0.85 by x4.1; the linear estimate 6.69 is outside the verified nef range, so the reported score uses the conservative verified-nef ceiling 1.893 as anchor - true N(E_F) may be higher (no saturation observed in-domain) or lower. Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 33.2; d=0.2222 -> 27.3; d=0.1667 -> 31.6; d=0.1111 -> 26.8. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially. This candidate's DOS lies beyond the calibration domain; the linear extrapolation is NOT used for the reported score (conservative ceiling anchor instead).",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2SNb2-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2SNb2-1",
|
| 4 |
+
"osc_id": "OSC-00247",
|
| 5 |
+
"formula": "Nb4S2",
|
| 6 |
+
"method_name": "Physics-screened discovery: Nb4S2 (2SNb2-1), calibrated N(E_F) ~ 1.04 states/eV/atom from C2DB PDOS. Non-magnetic d-metal, p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate). Conditional dome score 17.4 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 17.44,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 1.04; (4) e_hull=0.036 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 18.3; d=0.2222 -> 15.1; d=0.1667 -> 17.4; d=0.1111 -> 14.8. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially.",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|
submissions/Tuli99_discovery_2SeNb2-1.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"author": "Tuli99",
|
| 3 |
+
"material_id": "2SeNb2-1",
|
| 4 |
+
"osc_id": "OSC-00249",
|
| 5 |
+
"formula": "Nb4Se2",
|
| 6 |
+
"method_name": "Physics-screened discovery: Nb4Se2 (2SeNb2-1), calibrated N(E_F) ~ 1.15 states/eV/atom from C2DB PDOS. Non-magnetic d-metal, p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate). Conditional dome score 19.1 at delta=1/6; organizer downfolding requested.",
|
| 7 |
+
"provisional_score": 19.145,
|
| 8 |
+
"verified": false,
|
| 9 |
+
"track": "discovery",
|
| 10 |
+
"discovery_note": "Screening evidence (v3.1, full-pool DOS scan of 267 candidates): (1) hard filters from enrichment analysis of the verified active set - non-magnetic (0/30 top verified scorers are magnetic vs 28% base rate) and d-metal; (2) p4/nmm lattice, 4-fold symmetric (layer-group prior: 47% of the top-30 verified materials are 4-fold vs 13% base rate); (3) N(E_F) estimated from C2DB PDOS at E_F divided by n_atoms and the calibration median ratio 0.5198 (47 active-material samples re-measured with the fixed parser, mean |err| 0.186, max 0.606) -> 1.15; (4) e_hull=0.002 eV/atom (C2DB thermodynamic distance to the convex hull; 0 = on-hull stable). Conditional score table over the dome window (A_d from public scoring_meta): d=0.2778 -> 20.1; d=0.2222 -> 16.5; d=0.1667 -> 19.1; d=0.1111 -> 16.2. delta cannot be predicted from composition (active-set evidence: CuI2 0.1667 vs CuBr2 0.7222), hence the conditional form. A_d(delta) reference: our own ED engines, validated 27/27 against the published independent ED (reproducibility package in PR #59); at delta=1/6 our curve gives 0.056827 (+2.0% of the organizer anchor 0.05569).",
|
| 11 |
+
"hypothesis_only": true,
|
| 12 |
+
"estimated_score_source": "Calibrated C2DB PDOS heuristic for N(E_F) (median-ratio calibration, 47 active samples, mean |err| 0.19) + organizer's public A_d(delta) table; NOT a DFT/ED calculation for this proposed structure.",
|
| 13 |
+
"uncertainty": "Active-band filling (delta) depends on the organizer's SCF/downfolding and active-band selection and cannot be predicted from composition; the actual score can be zero if delta falls outside the dome. N(E_F) calibration error: mean 0.186 / max 0.606 states/eV/atom (47 samples). Polytype twins inside the verified set differ in DOS at E_F by up to ~3x (stacking sensitivity), so the same composition in another stacking can shift N(E_F) substantially.",
|
| 14 |
+
"at": 1792592070
|
| 15 |
+
}
|