# Causal Spatial Reasoning on CausalSpatial Level 1 Source: [Causal Spatial Reasoning on CausalSpatial Level 1]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:39:31.135Z Evaluation of model performance on the CausalSpatial Level 1 dataset, focusing on causal and spatial reasoning capabilities, measured by anticipation scores across collision, compatibility, occlusion, and trajectory prediction. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Causal Spatial Reasoning | | Dataset (including any stated split) | CausalSpatial Level 1 | | Metric | Collision Anticipation | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Collision Anticipation](): Higher is better; unit: PERCENT. - [Compatibility Anticipation](): Higher is better; unit: PERCENT. - [Occlusion Anticipation](): Higher is better; unit: PERCENT. - [Trajectory Anticipation](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. - [Not Sure Rate (NSR)](): Lower is better; unit: PERCENT. ## Recorded results | Method | Collision Anticipation | Result date | Report paper | | --- | --- | --- | --- | | Human | 86.84 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Sonnet 3.7 | 55 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-30B-A3B-T. | 52.99 | Not recorded | 69705ddae25ad2fa9d1ed67a | | GPT5 | 49.65 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-4B-I. | 47.55 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-2B-I. | 47.43 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-30B-A3B-I. | 47.02 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Gemini2.5 Flash | 46.21 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Gemini2.5 Pro | 43.62 | Not recorded | 69705ddae25ad2fa9d1ed67a | | LLaVAOneVision1.5-8B-I. | 43.01 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-8B-T. | 42.74 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Spatial-MLLM | 41.96 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen2.5-VL-7B-I. | 41.75 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Qwen3-VL-8B-I. | 41.4 | Not recorded | 69705ddae25ad2fa9d1ed67a | | GPT5-mini | 37.76 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Random | 23.61 | Not recorded | 69705ddae25ad2fa9d1ed67a | | Spatial-VLM | 20.98 | Not recorded | 69705ddae25ad2fa9d1ed67a | ### Human Collision Anticipation: 86.84. Result ID: 6997a7bf5698665ec4fddbdc. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Sonnet 3.7 Collision Anticipation: 55. Result ID: 6997a7bf5698665ec4fddbdb. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-VL-30B-A3B-T. Collision Anticipation: 52.99. Result ID: 6997a7bf5698665ec4fddbd6. [Reporting paper]() Recorded evaluation configuration: {"Scale":"30B-A3B","Reasoning\_Trained":true}. This record does not assert independent reproduction. ### GPT5 Collision Anticipation: 49.65. Result ID: 6997a7bf5698665ec4fddbd7. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen3-VL-4B-I. Collision Anticipation: 47.55. Result ID: 6997a7bf5698665ec4fddbd2. [Reporting paper]() Recorded evaluation configuration: {"Scale":"4B"}. This record does not assert independent reproduction. ### Qwen3-VL-2B-I. Collision Anticipation: 47.43. Result ID: 6997a7bf5698665ec4fddbd1. [Reporting paper]() Recorded evaluation configuration: {"Scale":"2B"}. This record does not assert independent reproduction. ### Qwen3-VL-30B-A3B-I. Collision Anticipation: 47.02. Result ID: 6997a7bf5698665ec4fddbd5. [Reporting paper]() Recorded evaluation configuration: {"Scale":"30B-A3B"}. This record does not assert independent reproduction. ### Gemini2.5 Flash Collision Anticipation: 46.21. Result ID: 6997a7bf5698665ec4fddbda. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini2.5 Pro Collision Anticipation: 43.62. Result ID: 6997a7bf5698665ec4fddbd9. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### LLaVAOneVision1.5-8B-I. Collision Anticipation: 43.01. Result ID: 6997a7bf5698665ec4fddbcf. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Scale":"8B"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-T. Collision Anticipation: 42.74. Result ID: 6997a7bf5698665ec4fddbd4. [Reporting paper]() Recorded evaluation configuration: {"Scale":"8B","Reasoning\_Trained":true}. This record does not assert independent reproduction. ### Spatial-MLLM Collision Anticipation: 41.96. Result ID: 6997a7bf5698665ec4fddbce. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-I. Collision Anticipation: 41.75. Result ID: 6997a7bf5698665ec4fddbd0. [Reporting paper]() Recorded evaluation configuration: {"Scale":"7B"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-I. Collision Anticipation: 41.4. Result ID: 6997a7bf5698665ec4fddbd3. [Reporting paper]() Recorded evaluation configuration: {"Scale":"8B"}. This record does not assert independent reproduction. ### GPT5-mini Collision Anticipation: 37.76. Result ID: 6997a7bf5698665ec4fddbd8. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Random Collision Anticipation: 23.61. Result ID: 6997a7bf5698665ec4fddbdd. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Spatial-VLM Collision Anticipation: 20.98. Result ID: 6997a7bf5698665ec4fddbcd. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.