1 무엇을 어떻게 쟀나
VLM eval(S3) 답 {category, room, surface}을 재생: 벤치마크 종료 상태(final_state.npz 25 pose + 로봇 base) 복원 → 답의 (room, surface)를 접지 v3(답 granularity까지 메모리, 그 가구 위 인스턴스 찾기는 하네스, 후보 ≤2) → 텔레포트 → Stage B reached(GT ≤0.85 m) → Stage C VLA(mixv2-20k, zed2 입력) "Pick up the {category}" 600스텝 × 조건부 재시도 2, 판정 GT body any-of lift(프로브 실행 전 생성). 집계는 질의 단위(기권·미접지·미도달 = 0), reached_only 질의는 Stage B가 최종.
접지 규칙 v1→v3 진화(oracle 0.63→0.82)와 Stage C 재시도·프로브 버그 수정(스모크에서 발견)은 claude/260821/exp-e2e_exec_stage_b.md, exp-e2e_exec_stage_c.md 참조.
2 셀별 SR_B / SR_C (질의 단위, n=gradable)
| cell | n | abstain | SR_B | SR_C | P(lift|reached, lifted-only) | n_lift_eval |
|---|
| oracle-loc | 348 | 0.00 | 0.830 | 0.239 | 0.166 | 247 |
| frames-uniform@gemini-3.6-flash | 348 | 0.03 | 0.589 | 0.178 | 0.192 | 177 |
| rollforward-vis-pure@gemini-3.6-flash | 348 | 0.02 | 0.612 | 0.167 | 0.171 | 187 |
| textlog@gemini-3.6-flash | 348 | 0.38 | 0.405 | 0.109 | 0.149 | 121 |
| textlog@claude-sonnet-5 | 348 | 0.41 | 0.374 | 0.101 | 0.144 | 111 |
| frames-uniform@claude-sonnet-5 | 348 | 0.05 | 0.362 | 0.095 | 0.184 | 114 |
| rollforward-vis-pure@claude-sonnet-5 | 348 | 0.05 | 0.379 | 0.095 | 0.161 | 118 |
| rollforward-v2@claude-sonnet-5 | 348 | 0.32 | 0.296 | 0.078 | 0.146 | 89 |
| blind@gemini-3.6-flash | 348 | 0.41 | 0.216 | 0.066 | 0.175 | 63 |
| rollforward-v2@gemini-3.6-flash | 348 | 0.37 | 0.244 | 0.066 | 0.162 | 74 |
| blind@claude-sonnet-5 | 348 | 0.53 | 0.152 | 0.032 | 0.125 | 48 |
플랜 단위 Stage C 성공(소스별): blind@claude-sonnet-5 6/12, blind@gemini-3.6-flash 11/16, frames-uniform@claude-sonnet-5 21/28, frames-uniform@gemini-3.6-flash 34/52, oracle-loc 41/69, rollforward-v2@claude-sonnet-5 13/19, rollforward-v2@gemini-3.6-flash 12/22, rollforward-vis-pure@claude-sonnet-5 19/29, rollforward-vis-pure@gemini-3.6-flash 32/48, textlog@claude-sonnet-5 16/32, textlog@gemini-3.6-flash 18/34
4 실패 사례 (oracle-loc 기준 — 메모리가 아니라 하네스/정책 한계)
val_11__2 · Pick up the remote control I moved earlier · sources oracle-loc, textlog@gemini-3.6-flash, frames-uniform@gemini-3.6-flash…
FAIL · 476s · sel0_a0(remotecontrol_dbbd): t0:✗600, t1:✗600
val_144__0 · Pick up the apple I moved earlier · sources oracle-loc, blind@gemini-3.6-flash, textlog@gemini-3.6-flash…
FAIL · 494s · sel0_a0(apple_ff93afa9f54b): t0:✗600, t1:✗600
val_179__0 · Pick up all the ladles · sources oracle-loc
FAIL · 662s · sel0_a0(ladle_efa098c0c2ee): t0:✓300 | sel1_a0(ladle_efa098c0c2ee): t0:✗600 | sel1_a1(ladle_efa098c0c2ee): t0:✗600
val_101__1 · Pick up all the bowls I moved earlier · sources oracle-loc
FAIL · 626s · sel0_a0(bowl_46a21212675e4): t0:✓189 | sel1_a0(painting_2fadc8a08): t0:✗600 | sel1_a1(bowl_6befd62f08fd3): t0:✗600
성공 사례
val_101__1 · Pick up the remote control I moved earlier · sources oracle-loc, textlog@gemini-3.6-flash, rollforward-v2@gemini-3.6-flash…
SUCCESS · 318s · sel0_a0(remotecontrol_dbbd): t0:✗600 | sel0_a1(remotecontrol_dbbd): t0:✓118
val_125__0 · Pick up the fork · sources oracle-loc, frames-uniform@gemini-3.6-flash, rollforward-vis-pure@gemini-3.6-flash…
SUCCESS · 25s · sel0_a0(fork_a17d780ca6a79): t0:✓55
val_179__0 · Pick up the tissue box · sources oracle-loc, frames-uniform@gemini-3.6-flash, rollforward-vis-pure@gemini-3.6-flash…
SUCCESS · 136s · sel0_a0(tissuepaper_1f2829): t0:✓327