Spaces:
Running
Running
Commit History
-
Update to uniform US spelling
2fd6a71
-
Add dynamic thumbnails for orgs
e068beb
-
Untrack accidentally-committed noise files; gitignore the patterns
The previous commit (b6dab21) pulled in several files that should not be tracked: - "(1).tsx" — Finder duplicates of route files - output/hierarchy_explorer.html — local probe output - public/peer-ranks.json — local-generated cache (~370k lines) - app/embed/eval/distribution/[...id]/page.tsx — orphan route Remove from the index and extend .gitignore so they can't sneak back through a `git add -A` or accidental wildcard. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>445ce35 -
view-data: paranoid CAST every output column to a primitive type
The previous patch left simple scalars (VARCHAR / BOOLEAN / DOUBLE / INTEGER) untouched, on the assumption the binding handles those natively. The deployed log still surfaces "Invalid Error: don't know what type:" on the new explicit-column SELECT, which means whatever the binding is choking on is reachable even without r.* expansion and even with every nested type wrapped. The only thing left to vary is whether the binding ever sees the parquet's native logical types at all. Force every output column through `CAST(... AS <primitive>)`. The binding now only ever sees VARCHAR / BOOLEAN / INTEGER / DOUBLE in the result schema regardless of what type the underlying parquet column declared. Local prod build keeps returning 200 on /api/model-summary with model_info correctly nested. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>b6dab21 -
Embeds: histogram route, leaderboard slices + sort, brand mark; cross-source row dedup
Embed routes - New /embed/eval/histogram/<evalId>?model=<routeId> renders a single model/benchmark plotbox inside <BenchmarkDetail> via a new embedSurface="histogram" + embedTargetEvalId; the on-page split picker, view selector, peer rank, cross-family whisker and setup notes all flow through unchanged. The plotbox EmbedButton points at this route with label "Histogram". - Leaderboard embed: SPLIT dropdown for subtask-sliced evals (Global MMLU style), click-to-sort column headers with direction toggle, and getMetricChipLabel fallback so cells stop collapsing to "Score" when display_name is generic. - Frontier embed: same slice picker, locked to frontier view (showViewToggle=false), shared metric-label fallback. - Embed layout: small EvalEval logo + text in the top-left of every embed, linking to evalcards.evalevalai.com. Cross-source row dedup - New dedupeLeaderboardRowsByModelIdentity in lib/eval-processing.ts collapses two-source duplicates (e.g. one row develo…7a54021
-
Cross-source dedup, plotbox polish, pretty URLs, eval page fallbacks
Model page (benchmark-detail.tsx) — cross-source grouping rework: - Category-mode tiles dedupe per benchmark canonical (8x MMLU-Pro across helm/vals-ai/openeval/… → 1 tile + cross-family whisker overlay), with an "N sources" badge. - Fixed benchmarkIndexLookup to prefer benchmark canonicals over umbrella entries (suite/aggregator) when building eval_summary_id → canonical_key. Without that, ids like artificial-analysis-llms/mmlu-pro mapped to "artificial analysis" instead of "mmlu-pro" and dedup leaked. - Aggregator-only score dedup: byte-equal scores from llm-stats are dropped (republished canonical numbers) but two independent evaluators arriving at the same number both stay — confirming signal, not duplicate data. Plus an eval_summary_id dedup before score dedup so the same observation listed under two family_keys in benchmark_index doesn't double-count. - Multi-category bucketing: a benchmark with curated tags ["general", "re…0b45710
-
Make /models tables column-sortable; rebalance /evals + /models toolbars
* /models, /developers/[id]: replace the 'Sort · X' select with click-to-sort column headers. Each column toggles direction on repeat click; switching columns picks a sensible default direction (asc for name/developer, desc for numeric / recency). Defaults land on Released ↓ for the model tables and Models ↓ for the developer table so the newest models surface first. * /developers (developer table): drop the 'Top evaluations' column — its chip strip duplicated info that lives one click deeper on the developer page and was crowding the row. * /evals: fold the standalone search row into the existing meta row (Families / Composites / Single benchmarks / Slices / Metrics) with the input pinned right via ml-auto. Same compaction on /models — search now sits at ml-auto where the sort dropdown used to be and widens to a 360px cap so it fills the freed space cleanly. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>5a2d59c
-
Tighten eval cards UI and clean up stale local data
UI changes - Replace the stand-alone reproducibility/per-row signal indicators with a single benchmark-level interpretive-signals strip (paper §4.2.1). Strip is rendered compactly inside the technical-details accordion, sized to match the Card Quality Notes box; large 4-tile layout was doing too much. Per-record reproducibility/provenance display in leaderboard rows is removed since signals are corpus/eval-level. - Theme the param range picker against the design system (square thumbs, hairline rail, mono ticks). Three variants (default / inline / promo). Visibility of the picker is gated on whether any model in the leaderboard has a detectable parameter count, with an explicit "Show models without known size" toggle when present. - Eval list: drop the near-empty Suites/Slices/Metrics/Organisation columns, drop the always-grey category dot, expose a one-line family description sourced from the navigation target (no leaf borrowing), and add a category pill selecto…32864b0 -
Add new component files and align app to EvalEval design system
Adds score-distribution, model-table, eval-node-table, family-table, developer-table, infinite-scroll components and logo assets that were missing from the repo, causing the build to fail. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>dbdd6d1
-
fix bugs
04b4cff
-
feat: refine model and benchmark exploration
03e2430