Open
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Issue #29 - Hide partial benchmark runs by default: - Add 'Hide partial runs' toggle to Task Heatmap view (default: checked) - Track isComplete flag for each model based on category coverage - Show count of hidden partial runs in heatmap description - Add proper empty state when all models are partial Issue #22 - Category selection filter for benchmark visualizations: - Add CategoryFilter component with multi-select dropdown - Filter tasks displayed in heatmap by selected categories - Recalculate model scores based on selected categories only - Persist category selection in URL params for shareability Shared infrastructure: - Add lib/categories.ts with utility functions for category detection - Add category grouping and filtered score calculation helpers This prevents misleading scores like '100%' from models that only completed a single category, and lets users focus on categories that matter most to them. Closes #22 Closes #29
olearycrew
force-pushed
the
feat/category-filters
branch
from
7b06b81 to
b0fe074
Compare
CATEGORY_ICONS only defined 10 categories (api, validation, calendar, etc.) but the API returns 15 different categories (basic, file_ops, creative, data_analysis, memory, organization, synthesis, content_transformation, etc.). This caused two bugs: 1. hasAllCategories() always returned false because it checked for categories like 'api' and 'validation' that don't exist in the data, so hidePartialRuns (default: true) filtered out ALL models -> 'No task data available' 2. Category filter with all selected still missed tasks in unrecognized categories (file_ops, creative, etc.) since they weren't in ALL_CATEGORIES Also fixed CategoryFilter label showing 'All categories' when none were selected (noneSelected case was incorrectly merged with allSelected).
olearycrew
requested changes
Mar 16, 2026
Member
olearycrew
left a comment
olearycrew
left a comment
There was a problem hiding this comment.
The task heatmap is now blank. Need that fixed first
hasAllCategories() checks against the full CATEGORY_ICONS list (15 categories), but older benchmark versions only had 7-9 categories. With hidePartialRuns defaulting to true, submissions from older benchmarks were incorrectly filtered out as 'partial' runs since they're missing categories that didn't exist in their version. Default to false until we track per-version category counts. URL param changed from ?partial=show (to show) to ?partial=hide (to hide).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
This PR implements two related issues for improving benchmark visualization accuracy:
Issue #29 - Hide partial benchmark runs by default
Issue #22 - Category selection filter
?categories=coding,api)Implementation Details
lib/categories.tswith utility functions:hasAllCategories()- check if submission is completecalculateFilteredScore()- recalculate scores for selected categoriesgroupTasksByCategory()- organize tasks by categoryCategoryFiltercomponent with dropdown UITaskHeatmapto support both filtersScreenshots
The filters appear in the Task Heatmap view under Graphs tab
Testing
npm run build)Closes #22
Closes #29