Public scan — anyone with this URL can view this analysis. Sign up to track your own repos privately, run scheduled re-scans, and get AI fix prompts via your dashboard.
Sign up free Scan another repo

SohrabTa/sparse-crosscoders-prott5

https://github.com/SohrabTa/sparse-crosscoders-prott5 · scanned 2026-06-16 04:39 UTC (1 month, 3 weeks ago)

20 raw signals (0 security + 20 graph)

File as issue Image
UNIFIED Repobility · multi-layer engine · AI coders

Complete repo analysis

Last scanned 1 month, 3 weeks ago · v1 · 17 actionable findings from 1 signal source. 3 repeated signals grouped for readability. Security checks, system graph analysis, and verified AI-agent feedback are merged into one review queue.

JSON
Combined
80.2
/100
Security checks
80
0 findings
System graph
80
55.6% cov · 17 findings
Critical
0
click to filter
Agents
10
0 votes
Crowd
0
reported
Severity distribution — click a segment to filter
Active filters: excluding tests × Reset all
Severity: Critical 0 High 0 Medium 7 Low 7 9-Layer: Software Security Quality Integrity Frontend Hardware Data Network Cicd Source: Security checks 0 System graph 17 Crowd 0 Layer: Software 2 Network 4 Security 1 Quality 10

All 164 nodes from the latest scan, grouped by kind. Each node is a unit the engine identified (file, function, endpoint, table…). Most users won't need this view — it's primarily for debugging the engine's graph extraction or for AI agents that want to enumerate the project structure.

LabelLayerStatusPath
read_msa software healthy harvest_msa_conservation.py:46
seq_weights software healthy harvest_msa_conservation.py:63
column_entropies software healthy harvest_msa_conservation.py:77
process_assay software healthy harvest_msa_conservation.py:92
encode software healthy harvest_msa_conservation.py:107
main software healthy harvest_msa_conservation.py:134
main software healthy per_residue_identity_check.py:47
main software healthy proteingym_category_specificity.py:78
run software healthy proteingym_category_specificity.py:107
main software healthy build_proteingym_matches.py:29
parse_concept_name software healthy match_proteingym_concepts.py:122
concept_name_for software healthy match_proteingym_concepts.py:132
load_pairings software healthy match_proteingym_concepts.py:140
parse_mutant_positions software healthy match_proteingym_concepts.py:192
fetch_uniprot_row software healthy match_proteingym_concepts.py:203
concept_to_residue_set software healthy match_proteingym_concepts.py:254
main software healthy match_proteingym_concepts.py:304
decoder_norms_per_layer software healthy plot_layerwise_feature_anatomy.py:27
best_pairing_per_feature software healthy plot_layerwise_feature_anatomy.py:37
main software healthy plot_layerwise_feature_anatomy.py:45
feats software healthy plot_layerwise_feature_anatomy.py:61
parse_mutant_positions software healthy harvest_raw_prott5.py:60
__init__ software healthy harvest_raw_prott5.py:77
encode software healthy harvest_raw_prott5.py:90
hook software dead harvest_raw_prott5.py:104
harvest_one_assay software healthy harvest_raw_prott5.py:122
main software healthy harvest_raw_prott5.py:221
count_proteins software healthy count_proteins.py:8
gate_stats_vs_batchtopk software healthy compare_jumprelu_modes.py:59
fvu_accumulate software healthy compare_jumprelu_modes.py:68
main software healthy compare_jumprelu_modes.py:111
batches software healthy compare_jumprelu_modes.py:144
filter_sequences software healthy create_dataset.py:7
sample_sequences software healthy create_dataset.py:46
create_dataset software healthy create_dataset.py:89
fig_f1_vs_fitness software healthy plot_proteingym_paper_figs.py:31
fig_chance_control software healthy plot_proteingym_paper_figs.py:55
fig_holdout software healthy plot_proteingym_paper_figs.py:79
main software healthy plot_proteingym_paper_figs.py:96
spearman software healthy diagnostic_ridge_probe.py:54
cv_ridge_spearman software healthy diagnostic_ridge_probe.py:66
diag_raw_one_assay software healthy diagnostic_ridge_probe.py:99
load_scoring_parquet software healthy diagnostic_ridge_probe.py:125
diag_sae_one_pair software healthy diagnostic_ridge_probe.py:130
main software healthy diagnostic_ridge_probe.py:159
is_point_site software healthy design1_concept_localized_fitness.py:97
auc_from_firing software healthy design1_concept_localized_fitness.py:101
auc_from_ranks software healthy design1_concept_localized_fitness.py:113
mask_perm_p software healthy design1_concept_localized_fitness.py:123
restricted_auc software healthy design1_concept_localized_fitness.py:152

Showing first 50 of this kind. Full payload available via the JSON button at the top of the page.

LabelLayerStatusPath
harvest_msa_conservation.py software healthy harvest_msa_conservation.py
submit_score.sh software healthy submit_score.sh
per_residue_identity_check.py software healthy per_residue_identity_check.py
proteingym_category_specificity.py software healthy proteingym_category_specificity.py
build_proteingym_matches.py software healthy build_proteingym_matches.py
match_proteingym_concepts.py software healthy match_proteingym_concepts.py
submit_pooled.sh software healthy submit_pooled.sh
plot_layerwise_feature_anatomy.py software healthy plot_layerwise_feature_anatomy.py
harvest_raw_prott5.py software healthy harvest_raw_prott5.py
count_proteins.py software healthy count_proteins.py
compare_jumprelu_modes.py software healthy compare_jumprelu_modes.py
create_dataset.py software healthy create_dataset.py
proteingym_category_counts.py software healthy proteingym_category_counts.py
plot_proteingym_paper_figs.py software healthy plot_proteingym_paper_figs.py
diagnostic_ridge_probe.py software healthy diagnostic_ridge_probe.py
README.md software healthy README.md
pyproject.toml software healthy pyproject.toml
design1_concept_localized_fitness.py software healthy design1_concept_localized_fitness.py
download_uniprot_dataset.py software healthy download_uniprot_dataset.py
qualitative_proteingym_analysis.py software healthy qualitative_proteingym_analysis.py
chance_control.py software healthy chance_control.py
verify_jumprelu_conversion.py software healthy verify_jumprelu_conversion.py
harvest_per_residue_wt.py software healthy harvest_per_residue_wt.py
convert_batchtopk_to_jumprelu.py software healthy convert_batchtopk_to_jumprelu.py
per_residue_and_holdout.py software healthy per_residue_and_holdout.py
position_residualized_spearman.py software healthy position_residualized_spearman.py
score_proteingym_features.py software healthy score_proteingym_features.py
analyze_proteingym_scoring.py software healthy analyze_proteingym_scoring.py
submit_full_feat.sh software healthy submit_full_feat.sh
migrate_gated_schema.py software healthy migrate_gated_schema.py
harvest_pooled_metrics.py software healthy harvest_pooled_metrics.py
harvest_full_feature_spearman.py software healthy harvest_full_feature_spearman.py
extract_activations.py software healthy extract_activations.py
submit_baseline_proteingym.sh software healthy submit_baseline_proteingym.sh

LabelLayerStatusPath
port:15 network healthy submit_score.sh
port:13 network healthy submit_pooled.sh
port:24 network healthy submit_full_feat.sh
port:14 network healthy submit_baseline_proteingym.sh

LabelLayerStatusPath
SohrabTa__sparse-crosscoders-prott5 software healthy /data/fable5_failed_archive/SohrabTa__sparse-crosscoders-pr…

LabelLayerStatusPath
ProtT5SingleLayerHarvester software healthy harvest_raw_prott5.py:70

LabelLayerStatusPath
gpu (detected) hardware healthy harvest_msa_conservation.py
For AI agents: Voting guide (TP/FP) MCP manifest Stdio wrapper SARIF Integrate Findings queue Vote TP/FP on findings to calibrate the engine.
For AI agents + API integrations
Email me when this repo regresses
Free. We re-scan periodically; new criticals → your inbox. No signup required for the scan itself.
API access

This page is publicly accessible at: https://repobility.com/scan/ca766c17-5a0e-4d2b-868b-838338da08ce/

To check status programmatically (no auth required): 

curl -s https://repobility.com/api/v1/public/scan/ca766c17-5a0e-4d2b-868b-838338da08ce/

Important — please don't re-submit the same URL repeatedly. The submission endpoint is idempotent: re-submitting the same git URL returns this same scan_token, not a new one. To re-scan this repo, sign up free and use the dashboard.