MolmoAct 2 Evaluation gallery
IN PROGRESS · PROVISIONAL Only fully completed, validated task shards are shown. Campaign coverage and aggregate rates can change.
Paper protocol Task 004 Validated task

pick up the black bowl in the top drawer of the wooden cabinet and place it on the plate

BDDL language: pick the akita black bowl in the top layer of the wooden cabinet and place it on the plate

Completed task success rate
94.0%
47 of 50
Control horizon
280
Maximum steps
Camera videos
100
Two per rollout
Evaluation time
7m 50s
Task process wall time

Canonical catalog

Task definition

Model instruction
pick up the black bowl in the top drawer of the wooden cabinet and place it on the plate
BDDL language
pick the akita black bowl in the top layer of the wooden cabinet and place it on the plate
Problem
libero_tabletop_manipulation
Objects
akita_black_bowl_1, akita_black_bowl_2, cookies_1, glazed_rim_porcelain_ramekin_1, plate_1
Fixtures
flat_stove_1, main_table, wooden_cabinet_1
Initial-state predicates
8
Fixed init states
50
Exact initial-state predicates 
(in akita_black_bowl_1 wooden_cabinet_1_top_region)
(on akita_black_bowl_2 wooden_cabinet_1_top_side)
(open wooden_cabinet_1_top_region)
(on plate_1 main_table_plate_region)
(on cookies_1 main_table_box_region)
(on glazed_rim_porcelain_ramekin_1 main_table_ramekin_region)
(on wooden_cabinet_1 main_table_cabinet_region)
(on flat_stove_1 main_table_stove_region)

Open exact BDDL definition

Simulator truth

Completion criteria

LIBERO checks these BDDL goal predicates after every action. All conditions must be true simultaneously in the same simulator step for the rollout to succeed.

  1. akita black bowl 1 is on plate 1

    (on akita_black_bowl_1 plate_1)
Exact parsed BDDL goal state 
(on akita_black_bowl_1 plate_1)

Selected rollout

LIBERO-Spatial · Task 004 · Episode 1

Success
Agent view observation.images.image
Eye in hand observation.images.image2
Videos are not loaded yet.

Rollout record

Episode metadata

Link to episode
Episode
1
Seed
1000
Init state
0
Control steps
128
Wall time
10s
Outcome
Success

Evidence browser

All rollouts

Select one episode to mount only its two camera streams in the player.

Episode Outcome Seed Init state Steps Video
1 Success 1000 0 128 119 KB
2 Success 1001 2 114 111 KB
3 Success 1002 4 134 121 KB
4 Success 1003 6 127 119 KB
5 Success 1004 8 120 113 KB
6 Success 1005 10 115 111 KB
7 Success 1006 12 134 120 KB
8 Success 1007 14 217 154 KB
9 Success 1008 16 121 111 KB
10 Success 1009 18 128 119 KB
11 Success 1010 20 126 117 KB
12 Success 1011 22 122 112 KB
13 Success 1012 24 123 114 KB
14 Success 1013 26 123 113 KB
15 Success 1014 28 122 114 KB
16 Success 1015 30 131 117 KB
17 Success 1016 32 119 119 KB
18 Success 1017 34 131 117 KB
19 Success 1018 36 139 129 KB
20 Failure 1019 38 280 233 KB
21 Success 1020 39 130 118 KB
22 Success 1021 41 143 124 KB
23 Success 1022 43 122 112 KB
24 Success 1023 45 115 111 KB
25 Success 1024 47 127 120 KB
26 Success 1025 49 116 113 KB
27 Success 1026 1 126 122 KB
28 Success 1027 3 126 114 KB
29 Success 1028 5 127 117 KB
30 Success 1029 7 124 119 KB
31 Success 1030 9 123 115 KB
32 Success 1031 11 129 124 KB
33 Success 1032 13 130 120 KB
34 Success 1033 15 121 115 KB
35 Success 1034 17 121 110 KB
36 Success 1035 19 124 116 KB
37 Success 1036 21 138 119 KB
38 Success 1037 23 132 118 KB
39 Success 1038 25 130 118 KB
40 Failure 1039 27 280 224 KB
41 Success 1040 28 132 119 KB
42 Success 1041 30 126 117 KB
43 Success 1042 32 111 107 KB
44 Success 1043 34 129 115 KB
45 Success 1044 36 129 121 KB
46 Success 1045 38 134 118 KB
47 Failure 1046 40 280 199 KB
48 Success 1047 41 134 125 KB
49 Success 1048 43 133 119 KB
50 Success 1049 45 121 113 KB

Integrity

Task attestation

Status
complete
Completed
Aug 16, 2026, 3:54 AM UTC
Manifest SHA-256
f30cb303ad…2ffd1614
Aggregate SHA-256
3f2c415d4a…15c515f6
Episode index SHA-256
a23b9767ab…423c4238
Validation SHA-256
60ed452146…52c978da
Artifact inventory SHA-256
40873c8bdc…0617e009