Overview
| Name | BEND · aliases: BEND, BEND 2.0 |
|---|---|
| Full name | Benchmarking DNA Language Models on Biologically Meaningful Tasks |
| Ambiguous names (not auto-linked) | bend |
| Family | Not a family entry. BEND 2.0 is treated as a new version of the same benchmark, not a separate family member. |
| Domain | Genomics & RNA |
| Inclusion | Supplementary entry |
| Type | B · Fixed datasets with frozen representations and a light downstream head. |
| Evidence level | L2 · Retrospective; the variant-effect tasks are zero-shot and nothing is blind. |
| Publisher | Technical University of Denmark, University of Copenhagen (Academic lab) |
| First release | · paper v1 Basis: arXiv 2311.12570 v1 source |
| Other milestones | |
| Latest update | · new version Basis: BEND 2.0 preprint (arXiv 2611.03806 v1) BEND 2.0 preprint released, reorganized around capabilities needed for DNA design: 8 tasks and 5 experimental datasets; the original repository's last commit was 2024-10-31. source The most recent sourced event, not a claim that nothing changed since. |
| Next expected | The authors have announced no date for a next version. |
| Status | ● Active · Version 2.0 released 2026-11-02; the original and 2.0 use different scopes and must be cited separately. |
| Scale | Original: 7 tasks, 285 to 2,005,617 samples, 512 to 100,096 bp. BEND 2.0: 8 tasks over 5 experimental datasets.
|
| Wet-lab refresh | Static data · Reuses published genome annotations and functional genomics data; no new experiments. |
| Own wet lab needed to evaluate a new model | No, run it as a download |
| Models reported | 0 counted |
| Unverified fields | † release_milestones.official_announcement, ranking.entries |
| As of | 2026-11-16 |
What it measures
- Task: Freeze pretrained embeddings and train a light head on human-genome functional tasks: gene finding, enhancer annotation, chromatin accessibility, histone modification, methylation and two variant-effect tasks.
- Truth: Published genome annotations and functional genomics measurements, with public labels for the variant-effect tasks.
- Primary metric: Per task: MCC, AUPRC or AUROC.
- Secondary metrics: BEND 2.0 adds correlation and design-task metrics.
- Version traps: The original and 2.0 define tasks differently and their numbers cannot be mixed; state the embedding chunk length, whether the official head configuration is used, and whether a readout is native or fitted.
Current ranking
Board
| Metric | Per-task metrics with no single official ranking metric: conditional log loss (bits/base, lower better), macro Spearman correlation, top-1 tissue identification accuracy |
|---|---|
| Eval set | BEND 2.0 Table 1 snapshot: 8 endpoints across 5 capability categories, 17 model-interface rows |
| Snapshot | |
| Board | https://machinelearninglifescience.github.io/BEND2 |
| Retrieved | 2026-11-16 |
Caveat. Values belong to BEND 2.0 (a 2026-11 preprint) and come from its site data file's Table 1 snapshot; there is no official aggregate, ranks hold per endpoint only, and native and fitted readouts are separate rows.
Shown in the overview table: Board
| Rank | Model | Score (verbatim) | Metric · eval set / protocol | Source type · date | Source and note |
|---|---|---|---|---|---|
| 1 | SpliceAI | 0.674 | Per-task metrics with no single official ranking metric: conditional log loss (bits/base, lower better), macro Spearman correlation, top-1 tissue identification accuracy BEND 2.0 Table 1 snapshot: 8 endpoints across 5 capability categories, 17 model-interface rows | Board | source Not a site model; first on this endpoint, as a specialist model's native output. |
| 2 | AlphaGenome | 0.650 | Per-task metrics with no single official ranking metric: conditional log loss (bits/base, lower better), macro Spearman correlation, top-1 tissue identification accuracy BEND 2.0 Table 1 snapshot: 8 endpoints across 5 capability categories, 17 model-interface rows | Board | source Second on this endpoint with a fitted readout and 0.621 native; first on five of the eight endpoints. |
| 4 | NTv3 650M post | 0.564 | Per-task metrics with no single official ranking metric: conditional log loss (bits/base, lower better), macro Spearman correlation, top-1 tissue identification accuracy BEND 2.0 Table 1 snapshot: 8 endpoints across 5 capability categories, 17 model-interface rows | Board | source Fourth on this endpoint; first on the sequence-prior endpoint at 1.375 bits/base, where lower is better. |
| 9 | Evo2 20B | 0.367 | Per-task metrics with no single official ranking metric: conditional log loss (bits/base, lower better), macro Spearman correlation, top-1 tissue identification accuracy BEND 2.0 Table 1 snapshot: 8 endpoints across 5 capability categories, 17 model-interface rows | Board | source Ninth on this endpoint; prior loss 1.388 bits/base (rank 3); last on the design endpoints. |
Models reported
Grouped by relation type; each line shows its source type. Counted relations add up to 0. A pairing whose evidence rests only on this survey's own page is shown as unverified.
None. No model among the 112 has a recorded pairing with this entry.
Caveats
- Human genome only, and frozen-embedding results are not a fine-tuning ceiling.
- The enhancer task is small and sparse and all models do poorly; text and tables disagree on histone modification.
- The original baselines are mostly models from around 2023, so newer DNA models are absent from that board.
- The two versions differ in scope: 2.0 is organized around design capabilities and its numbers cannot sit beside the original's.
- The site models on the 2.0 board are not yet recorded as pairings; their relation type awaits the next round.
How to use it
Take data and configs from the official repository, extract frozen embeddings per task and train a small head, which costs far less than fine-tuning. State the embedding chunk length and whether the official head configuration is used; report both mean rank and per-task values rather than a single task; keep the original and 2.0 apart; and report the zero-shot variant tasks alongside TraitGym.
Pairs well with: GUE + Nucleotide Transformer benchmark · Basenji/Enformer test set · TraitGym · ClinVar hold-out protocol
Sources
- BEND paper (arXiv 2311.12570 v1) (preprint) ·
- BEND paper full text v4 (preprint) ·
- ICLR 2024 conference paper page (paper)
- Official repository frederikkemarin/BEND (repository)
- BEND 2.0 preprint (arXiv 2611.03806) (preprint) ·
- BEND 2.0 leaderboard page (leaderboard) ·
Corrections
1 correction
| Field | Was | Now | Basis | Found |
|---|---|---|---|---|
| ranking.eval_set | original BEND standing | BEND 2.0 Table 1 snapshot (2026-10-22), per endpoint | source | 2026-11-16 |
Part of the AI for Bio benchmarks survey · all 44 benchmarks · open in the explorer. Record as of 2026-11-16.