Span-level detection on Social Media Posts ILE Detection
79.1PrecisionGPT-5.4 (No Taxonomy)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-5.4 (No Taxonomy)Model=GPT-5.4, Variant=No Taxonomy2026.06 | 79.1 | 48.3 | 59.9 | |
| GPT-5.4 (Calhoun & Fawcett)Model=GPT-5.4, Variant=Calhoun & Fawcett2026.06 | 78.7 | 47 | 58.9 | |
| Sonnet 4.6 (Calhoun & Fawcett)Model=Sonnet 4.6, Variant=Calhoun & Fawcett2026.06 | 77.8 | 56.4 | 65.4 | |
| Sonnet 4.6 (Zhang et al.)Model=Sonnet 4.6, Variant=Zhang et al.2026.06 | 77.1 | 53.3 | 63 | |
| Sonnet 4.6 (Leal-Arenas & Corizzo)Model=Sonnet 4.6, Variant=Leal-Arenas & Corizzo2026.06 | 77.1 | 61.9 | 68.6 | |
| GPT-5.4 (Zhang et al.)Model=GPT-5.4, Variant=Zhang et al.2026.06 | 76.4 | 45.3 | 56.9 | |
| GPT-5.4 (Leal-Arenas & Corizzo)Model=GPT-5.4, Variant=Leal-Arenas & Corizzo2026.06 | 76.4 | 55.1 | 64 | |
| Sonnet 4.6 (No Taxonomy)Model=Sonnet 4.6, Variant=No Taxonomy2026.06 | 74.6 | 61.8 | 67.6 | |
| DeepSeek V4 (Calhoun & Fawcett)Model=DeepSeek V4, Variant=Calhoun & Fawcett2026.06 | 74.2 | 50.6 | 60.2 | |
| GPT-5.4 (Fillies & Paschke)Model=GPT-5.4, Variant=Fillies & Paschke2026.06 | 74 | 53.5 | 62.1 | |
| Sonnet 4.6 (Fillies & Paschke)Model=Sonnet 4.6, Variant=Fillies & Paschke2026.06 | 72.8 | 62.4 | 67.2 | |
| DeepSeek V4 (Zhang et al.)Model=DeepSeek V4, Variant=Zhang et al.2026.06 | 72.3 | 44.8 | 55.3 | |
| Sonnet 4.6 (Proposed Taxonomy)Model=Sonnet 4.6, Variant=Proposed Taxonomy2026.06 | 72 | 66 | 68.8 | |
| DeepSeek V4 (Leal-Arenas & Corizzo)Model=DeepSeek V4, Variant=Leal-Arenas & Corizzo2026.06 | 72 | 56.3 | 63.2 | |
| DeepSeek V4 (Proposed Taxonomy)Model=DeepSeek V4, Variant=Proposed Taxonomy2026.06 | 71.3 | 58.9 | 64.5 | |
| GPT-5.4 (Proposed Taxonomy)Model=GPT-5.4, Variant=Proposed Taxonomy2026.06 | 70.7 | 62.3 | 66.2 | |
| DeepSeek V4 (Fillies & Paschke)Model=DeepSeek V4, Variant=Fillies & Paschke2026.06 | 69.3 | 53.6 | 60.5 | |
| DeepSeek V4 (No Taxonomy)Model=DeepSeek V4, Variant=No Taxonomy2026.06 | 68.3 | 53.4 | 59.9 |