Are submitted CIF structures currently used for Li-migration (E_a) scoring?

#1
by Ku-ku - opened

Hello,

A question about the Open Materials Challenge scoring.

The submit form says that attaching a CIF makes scoring faster and more accurate. In the current public leaderboard (fetched today, after the restart), the most recent row with an E_a value was scored on Sept 28; rows scored after that have E_a empty, and their notes still ask for a CIF.

  1. Does the current scoring worker read a CIF submitted in the structure field?
  2. Is the Li-migration (E_a) step running at the moment, or is it intentionally disabled?

Thank you.
β€” Ku-ku

FINAL_Bench org

짧은 λ‹΅μž…λ‹ˆλ‹€. 제좜 κ΅¬μ‘°λŠ” 쓰이지 μ•Šκ³  μžˆμ—ˆκ³ , μ§€κΈˆμ€ μ“°μž…λ‹ˆλ‹€.

λ¨Όμ € μš©μ–΄λ₯Ό λ§žμΆ”κ² μŠ΅λ‹ˆλ‹€. 이동μž₯벽은 점수 좕이 μ•„λ‹ˆλΌ κ΄€λ¬Έμž…λ‹ˆλ‹€. 리튬이 κ΄€ν†΅ν•˜λŠ” κ²½λ‘œκ°€ μžˆλŠ”μ§€, κ·Έ μž₯벽이 κΈ°μ€€ μ•ˆμ— λ“œλŠ”μ§€λ§Œ λ΄…λ‹ˆλ‹€. ν†΅κ³Όν•˜λ©΄ μˆœμœ„λŠ” μ‚°ν™” μ•ˆμ •μ„±, λ¦¬νŠ¬κΈˆμ† μ•ˆμ •μ„±, μš©λ„ μ‹ κ·œμ„± μ„Έ μΆ•μœΌλ‘œ μ •ν•©λ‹ˆλ‹€. κ·Έλž˜μ„œ μˆœμœ„ν‘œμ—λŠ” μž₯λ²½ μˆ˜μΉ˜κ°€ μ•„λ‹ˆλΌ κ΄€λ¬Έ 결과만 ν‘œμ‹œλ©λ‹ˆλ‹€.

  1. μ²¨λΆ€ν•˜μ‹  CIF κ°€ 이동μž₯λ²½ 계산에 μ“°μ˜€λŠ”μ§€: μ•„λ‹ˆμ—ˆμŠ΅λ‹ˆλ‹€. 제좜 κΈ°λ‘μ—λŠ” ꡬ쑰가 μ €μž₯돼 μžˆμ—ˆμœΌλ‚˜ μ±„μ κΈ°λ‘œ μ „λ‹¬ν•˜λŠ” κ²½λ‘œκ°€ μ—†μ—ˆμŠ΅λ‹ˆλ‹€.
  2. 이동μž₯λ²½ 단계가 돌고 μžˆμ—ˆλŠ”μ§€, μ˜λ„μ μœΌλ‘œ 끈 것인지: 끈 것이 μ•„λ‹™λ‹ˆλ‹€. 계산 λͺ¨λ“ˆ ν•˜λ‚˜κ°€ 채점 ν™˜κ²½μ— λΉ μ Έ μžˆμ–΄ 단계 전체가 μ‹€νŒ¨ν•˜κ³  μžˆμ—ˆμŠ΅λ‹ˆλ‹€.

μ„Έ κ°€μ§€λ₯Ό κ³ μ³€μŠ΅λ‹ˆλ‹€. λͺ¨λ“ˆμ„ μ±„μš°κ³ , 제좜 ꡬ쑰λ₯Ό μ±„μ κΈ°λ‘œ μ „λ‹¬ν•˜κ³ , ꡬ쑰 μš°μ„ μˆœμœ„λ₯Ό κΈ°μ‘΄ μˆ˜λ‘κ°’ λ‹€μŒμ— 첨뢀 ꡬ쑰, κ·Έ λ‹€μŒμ— 생성 ꡬ쑰둜 λ‘μ—ˆμŠ΅λ‹ˆλ‹€.

Ku-ku λ‹˜ 두 건은 10μ›” 25일에 μ²¨λΆ€ν•˜μ‹  ꡬ쑰둜 λ‹€μ‹œ μ±„μ ν–ˆμŠ΅λ‹ˆλ‹€. μˆœμœ„ν‘œμ˜ 채점 μ‹œκ°μœΌλ‘œ ν™•μΈν•˜μ‹€ 수 μžˆμŠ΅λ‹ˆλ‹€.

후보 이동μž₯λ²½ (eV) κ΄€λ¬Έ
OMC-C546AC9F3A 0.459 톡과
OMC-8EAA173E54 0.4688 톡과

이 두 μˆ˜μΉ˜λŠ” ν˜„μž¬ μˆœμœ„ν‘œμ— ν‘œμ‹œλ˜μ§€ μ•ŠμŠ΅λ‹ˆλ‹€. κ΄€λ¬Έ 결과만 λ‚¨κΈ°λŠ” μ„€κ³„μ˜€μŠ΅λ‹ˆλ‹€. μž₯λ²½ κ°’κ³Ό μ–΄λŠ ꡬ쑰둜 κ³„μ‚°ν–ˆλŠ”μ§€λ₯Ό ν•¨κ»˜ 보이도둝 λ°”κΎΈκ² μŠ΅λ‹ˆλ‹€. κ·Έ 편이 κ²°κ³Όλ₯Ό κ²€μ¦ν•˜κΈ° μ‰½μŠ΅λ‹ˆλ‹€.

ꡬ쑰λ₯Ό μ²¨λΆ€ν•œ λ‹€λ₯Έ μ œμΆœλ„ 같은 λ°©μ‹μœΌλ‘œ λ‹€μ‹œ 채점할 수 μžˆμŠ΅λ‹ˆλ‹€. μˆœμœ„ 전체가 μ›€μ§μ΄λŠ” λ²”μœ„λΌ 일정을 μ •ν•΄ κ³΅μ§€ν•˜κ² μŠ΅λ‹ˆλ‹€.


Short answer: submitted structures were not being used. They are now.

One clarification first. The migration barrier is a gate, not a scoring axis. It checks whether a percolating Li path exists and whether its barrier is within range. Candidates that pass are then ranked on oxidation stability, Li metal stability and application novelty. That is why the board shows a gate result rather than a barrier value.

  1. Was your attached CIF used for the barrier? No. The structure was stored with the submission, but no path carried it to the scorer.
  2. Was the barrier step running, or disabled on purpose? Not disabled. One calculation module was missing from the scoring environment, so the whole step failed.

Three fixes: the module is installed, submitted structures are passed through, and the structure order is now reference value, then your attached structure, then the generated one.

Your two candidates were rescored from your own structures on 25 October, which you can confirm from the scored time on the board. OMC-C546AC9F3A came out at 0.459 eV and OMC-8EAA173E54 at 0.4688 eV, both passing the gate.

Those values are not currently displayed, since the design kept only the gate result. We will change it to show the barrier and which structure it came from, which makes the result easier to check.

Other submissions with attached structures can be rescored the same way. That moves the whole ranking, so we will announce a schedule.

Sign up or log in to comment