Configuration

ESM-1b (per AA)

Protein language model embedding with the stated pooling, fed to a supervised head by the FLIP authors.

Overview

Protein language model embedding with the stated pooling, fed to a supervised head by the FLIP authors.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluations and results

Release 2026-09-17-134cd1815de8 · 14 evaluations · 14 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.

Evaluation setupMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
ESM-1b (per AA) on FLIP AAV-1-VS-REST: AAV fitness prediction, 1-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP AAV-1-VS-REST: AAV fitness prediction, 1-vs-rest splitDataset subset: FLIP AAV, 1-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.03 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 5, row(ESM-1b (per AA)), column(1-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP AAV-2-VS-REST: AAV fitness prediction, 2-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP AAV-2-VS-REST: AAV fitness prediction, 2-vs-rest splitDataset subset: FLIP AAV, 2-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.65 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 5, row(ESM-1b (per AA)), column(2-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP AAV-7-VS-REST: AAV fitness prediction, 7-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP AAV-7-VS-REST: AAV fitness prediction, 7-vs-rest splitDataset subset: FLIP AAV, 7-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.65 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 5, row(ESM-1b (per AA)), column(7-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP AAV-LOW-VS-HIGH: AAV fitness prediction, low-vs-high split Configuration: ESM-1b (per AA)Task: FLIP AAV-LOW-VS-HIGH: AAV fitness prediction, low-vs-high splitDataset subset: FLIP AAV, low-vs-high split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.39 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 5, row(ESM-1b (per AA)), column(low-vs-high) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP AAV-MUT-DES: AAV fitness prediction, Mut-Des split Configuration: ESM-1b (per AA)Task: FLIP AAV-MUT-DES: AAV fitness prediction, Mut-Des splitDataset subset: FLIP AAV, Mut-Des split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.76 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 5, row(ESM-1b (per AA)), column(Mut-Des) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP GB1-1-VS-REST: GB1 fitness prediction, 1-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP GB1-1-VS-REST: GB1 fitness prediction, 1-vs-rest splitDataset subset: FLIP GB1, 1-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.28 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 4, row(ESM-1b (per AA)), column(1-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP GB1-2-VS-REST: GB1 fitness prediction, 2-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP GB1-2-VS-REST: GB1 fitness prediction, 2-vs-rest splitDataset subset: FLIP GB1, 2-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.55 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 4, row(ESM-1b (per AA)), column(2-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP GB1-3-VS-REST: GB1 fitness prediction, 3-vs-rest split Configuration: ESM-1b (per AA)Task: FLIP GB1-3-VS-REST: GB1 fitness prediction, 3-vs-rest splitDataset subset: FLIP GB1, 3-vs-rest split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.79 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 4, row(ESM-1b (per AA)), column(3-vs-rest) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP GB1-LOW-VS-HIGH: GB1 fitness prediction, low-vs-high split Configuration: ESM-1b (per AA)Task: FLIP GB1-LOW-VS-HIGH: GB1 fitness prediction, low-vs-high splitDataset subset: FLIP GB1, low-vs-high split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.59 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 4, row(ESM-1b (per AA)), column(low-vs-high) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP RANDOM-SAMPLED-SPLITS-AAV: Random sampled splits fitness prediction, AAV split Configuration: ESM-1b (per AA)Task: FLIP RANDOM-SAMPLED-SPLITS-AAV: Random sampled splits fitness prediction, AAV splitDataset subset: FLIP Random sampled splits, AAV split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. The authors report these separately and call them optimistic: a random split of a mutational landscape leaks close relatives between train and test. Author-reported evaluation · Evaluation metadata: source checked
0.90 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 7, row(ESM-1b (per AA)), column(AAV) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP RANDOM-SAMPLED-SPLITS-GB1: Random sampled splits fitness prediction, GB1 split Configuration: ESM-1b (per AA)Task: FLIP RANDOM-SAMPLED-SPLITS-GB1: Random sampled splits fitness prediction, GB1 splitDataset subset: FLIP Random sampled splits, GB1 split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. The authors report these separately and call them optimistic: a random split of a mutational landscape leaks close relatives between train and test. Author-reported evaluation · Evaluation metadata: source checked
0.92 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 7, row(ESM-1b (per AA)), column(GB1) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP THERMOSTABILITY-HUMAN-CELL: Thermostability fitness prediction, Human-Cell split Configuration: ESM-1b (per AA)Task: FLIP THERMOSTABILITY-HUMAN-CELL: Thermostability fitness prediction, Human-Cell splitDataset subset: FLIP Thermostability, Human-Cell split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.76 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 6, row(ESM-1b (per AA)), column(Human-Cell) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP THERMOSTABILITY-HUMAN: Thermostability fitness prediction, Human split Configuration: ESM-1b (per AA)Task: FLIP THERMOSTABILITY-HUMAN: Thermostability fitness prediction, Human splitDataset subset: FLIP Thermostability, Human split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.71 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 6, row(ESM-1b (per AA)), column(Human) Source checking is not independent reproduction.
ESM-1b (per AA) on FLIP THERMOSTABILITY-MIXED: Thermostability fitness prediction, Mixed split Configuration: ESM-1b (per AA)Task: FLIP THERMOSTABILITY-MIXED: Thermostability fitness prediction, Mixed splitDataset subset: FLIP Thermostability, Mixed split (FLIP split) Trained on the split's training set with ten percent held out for validation, scored on its test set. Author-reported evaluation · Evaluation metadata: source checked
0.68 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedflip primary benchmark evidence · Table 6, row(ESM-1b (per AA)), column(Mixed) Source checking is not independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters