← all runs · ← report

wi007_qwen3_vl_8b_instruct_wi006 / live_d2_harness_pending_s408 not valid

What this trial was

runwi007_qwen3_vl_8b_instruct_wi006triallive_d2_harness_pending_s408conditionlivescenariopending_effectskinnotificationcanvas320×240 px (target 24 px)event at500 msresponse window16 sframe period16.67 msobservation publication delay30 msdeliverysimulated_fixed_delay, 20 msclockmonotonic_nseffect_delay_ns2000000000irrelevant_motionFalse

Controller: {"enable_thinking": false, "image_scale": 2, "invoke_on": "frame_change", "kind": "vlm_pending", "max_tokens": 64, "memory": "harness_pending", "model_id": "Qwen/Qwen3-VL-8B-Instruct", "pending_radius_px": 10, "pending_timeout_ns": 10000000000, "prompt_version": 4, "reasoning_effort": null, "temperature": 0.0}

Result for this trial

first click acceptedFalsefirst E→A—duplicates0invalid clicks3suppressed proposals0waits20decisions23exact successFalsedropped0contract violationsnone

Replay

key frames

Key frames: each decision's starting frame with the click it produced (amber = proposal before binding, red = committed, grey = suppressed), then the screen at each arrival A (green = valid).

The prompt sent with every frame

You control a desktop by clicking. You are shown one screenshot. At some point a notification card with a blue button appears at an unknown position and expires shortly after. If the card is visible, answer with action "click" and the coordinates of the center of its blue button on a 0-1000 grid (x: 0 = left edge, 1000 = right edge; y: 0 = top edge, 1000 = bottom edge). If no card is visible, answer with action "wait". Answer with the JSON object only.

[user] Screenshot attached. Click the target's center or wait.

Decision by decision — 23 decisions

#Bmodel done
B→D
model response (raw)parsedaction (canvas px)D (commit)A (receipt)out tokens
1531 ms1,068 ms
0.54 s
{"action": "wait"}
wait—
——7
21,081 ms1,644 ms
0.56 s
{"action": "wait"}
wait—
——7
31,658 ms2,279 ms
0.62 s
{"action": "wait"}
wait—
——7
42,294 ms2,914 ms
0.62 s
{"action": "wait"}
wait—
——7
52,930 ms3,550 ms
0.62 s
{"action": "wait"}
wait—
——7
63,565 ms4,190 ms
0.62 s
{"action": "wait"}
wait—
——7
74,205 ms4,828 ms
0.62 s
{"action": "wait"}
wait—
——7
84,843 ms5,402 ms
0.56 s
{"action": "wait"}
wait—
——7
95,416 ms6,040 ms
0.62 s
{"action": "wait"}
wait—
——7
106,056 ms6,677 ms
0.62 s
{"action": "wait"}
wait—
——7
116,693 ms7,316 ms
0.62 s
{"action": "wait"}
wait—
——7
127,332 ms7,956 ms
0.62 s
{"action": "wait"}
wait—
——7
137,971 ms8,595 ms
0.62 s
{"action": "wait"}
wait—
——7
148,610 ms9,231 ms
0.62 s
{"action": "wait"}
wait—
——7
159,247 ms9,806 ms
0.56 s
{"action": "wait"}
wait—
——7
169,820 ms10,441 ms
0.62 s
{"action": "wait"}
wait—
——7
1710,456 ms11,078 ms
0.62 s
{"action": "wait"}
wait—
——7
1811,093 ms11,718 ms
0.62 s
{"action": "wait"}
wait—
——7
1911,734 ms12,357 ms
0.62 s
{"action": "wait"}
wait—
——7
2012,372 ms14,135 ms
1.76 s
{"action": "click", "x": 687, "y": 484}
click
img (687, 484)
{'kind': 'click', 'x': 219, 'y': 116}
14,135 ms14,155 ms · invalid23
2114,181 ms15,922 ms
1.74 s
{"action": "click", "x": 604, "y": 374}
click
img (604, 374)
{'kind': 'click', 'x': 193, 'y': 89}
15,922 ms15,942 ms · invalid23
2215,965 ms17,708 ms
1.74 s
{"action": "click", "x": 687, "y": 484}
click
img (687, 484)
{'kind': 'click', 'x': 219, 'y': 116}
17,708 ms17,728 ms · invalid23
2317,731 ms18,283 ms
0.55 s
{"action": "wait"}
wait—
——7

Events and observations

ideventtask timeharness time
1target_appears500 ms500 ms
2window_closes16,500 ms16,500 ms

1020 observations published; first frame depicting the target published at 530 ms. Raw trace: live_d2_harness_pending_s408.jsonl.