Primary source: https://arxiv.org/html/2407.10362v3
Verified: 2026-10-12
Appendix B Table2; full public+private set, mean over3 model runs.
Human baseline tools allowed; models no tools. Human values from same table, not a language model.
L397: Table 2: Accuracy for all tasks/subtasks of LAB-Bench. Results are measured over the entire dataset (both public and private splits), and averaged over three runs per model.
L398: | Human | claude-3-5-sonnet-20240620 | claude-3-opus-20240229 | gemini-1.5-pro-001 | gpt-4o | gpt-4-turbo | claude-3-haiku-20240307 | Meta-Llama-3-70B-Instruct
L399: --- | --- | --- | --- | --- | --- | --- | --- | ---
L400: LitQA2 | 0.70 | 0.06 | 0.06 | 0.10 | 0.27 | 0.12 | 0.27 | 0.35
L401: SuppQA | 0.85 | 0.02 | 0.02 | 0.01 | 0.13 | 0.06 | 0.18 | 0.20
L402: FigQA | 0.77 | 0.46 | 0.24 | 0.25 | 0.29 | 0.25 | 0.23 | 0.00
L403: TableQA | 0.84 | 0.83 | 0.67 | 0.59 | 0.71 | 0.54 | 0.49 | 0.00
L404: DbQA_dga_task-v1 | 0.67 | 0.00 | 0.00 | 0.00 | 0.16 | 0.05 | 0.15 | 0.25
L405: DbQA_gene_location_task-v1 | 0.88 | 0.00 | 0.00 | 0.03 | 0.31 | 0.12 | 0.18 | 0.20
L406: DbQA_mirna_targets_task-v1 | 0.82 | 0.00 | 0.00 | 0.00 | 0.10 | 0.01 | 0.15 | 0.29
L407: DbQA_mouse_tumor_gene_sets-v1 | 1.00 | 0.55 | 0.36 | 0.10 | 0.57 | 0.16 | 0.48 | 0.53
L408: DbQA_oncogenic_signatures_task-v1 | 1.00 | 0.14 | 0.12 | 0.04 | 0.25 | 0.06 | 0.24 | 0.20
L409: DbQA_tfbs_GTRD_task-v1 | 0.88 | 0.00 | 0.00 | 0.00 | 0.01 | 0.00 | 0.11 | 0.19
L410: DbQA_variant_from_sequence_task-v1 | 0.60 | 0.02 | 0.00 | 0.00 | 0.42 | 0.00 | 0.22 | 0.30
L411: DbQA_variant_multi_sequence_task-v1 | 0.23 | 0.09 | 0.05 | 0.00 | 0.06 | 0.00 | 0.14 | 0.07
L412: DbQA_vax_response_task-v1 | 0.91 | 0.21 | 0.14 | 0.13 | 0.33 | 0.06 | 0.32 | 0.15
L413: DbQA_viral_ppi_task-v1 | 0.90 | 0.00 | 0.00 | 0.00 | 0.26 | 0.02 | 0.27 | 0.40
L414: ProtocolQA | 0.79 | 0.48 | 0.52 | 0.47 | 0.53 | 0.37 | 0.45 | 0.44
L415: SeqQA_ORF-seq-AAid-v1 | 0.91 | 0.07 | 0.03 | 0.02 | 0.16 | 0.07 | 0.17 | 0.19
L416: SeqQA_ORF-seq-AAseq-v1 | 0.88 | 0.87 | 0.67 | 0.41 | 0.50 | 0.39 | 0.57 | 0.49
L417: SeqQA_ORF-seq-numlen-v1 | 0.31 | 0.00 | 0.07 | 0.05 | 0.11 | 0.07 | 0.26 | 0.23
L418: SeqQA_ORF-transeff-v1 | 0.38 | 0.87 | 0.64 | 0.47 | 0.48 | 0.29 | 0.47 | 0.45
L419: SeqQA_PCR-gene-enzprimers-v1 | 0.83 | 0.57 | 0.75 | 0.47 | 0.60 | 0.49 | 0.61 | 0.58
L420: SeqQA_PCR-gene-gibshindprimers-v1 | 0.71 | 0.61 | 0.48 | 0.43 | 0.64 | 0.13 | 0.57 | 0.45
L421: SeqQA_PCR-gene-gibssmaprimers-v1 | 0.73 | 0.65 | 0.34 | 0.16 | 0.45 | 0.29 | 0.51 | 0.58
L422: SeqQA_PCR-geneprimers-enz-v1 | 0.94 | 0.73 | 0.76 | 0.66 | 0.74 | 0.66 | 0.83 | 0.37
L423: SeqQA_PCR-len-primers-v1 | 0.90 | 0.29 | 0.37 | 0.02 | 0.12 | 0.10 | 0.13 | 0.06
L424: SeqQA_PCR-primers-len-v1 | 0.90 | 0.35 | 0.29 | 0.15 | 0.13 | 0.19 | 0.26 | 0.21
L425: SeqQA_PCR-seq-enzprimers-v1 | 0.81 | 0.94 | 0.80 | 0.54 | 0.83 | 0.67 | 0.86 | 0.73
L426: SeqQA_PCR-seq-primers-v1 | 1.00 | 0.97 | 0.83 | 0.48 | 0.73 | 0.52 | 0.81 | 0.47
L427: SeqQA_Prop-seq-gcpercent-v1 | 1.00 | 0.59 | 0.45 | 0.39 | 0.29 | 0.18 | 0.39 | 0.35
L428: SeqQA_RE-seq-lenfrags-v1 | 0.83 | 0.12 | 0.13 | 0.07 | 0.17 | 0.07 | 0.21 | 0.19
L429: SeqQA_RE-seq-numfrags-v1 | 0.71 | 0.10 | 0.13 | 0.17 | 0.09 | 0.12 | 0.12 | 0.15
L430: CloningScenarios | 0.60 | 0.28 | 0.27 | 0.10 | 0.28 | 0.09 | 0.26 | 0.24