All pull requests

Pull requests

• dashboard: multi-objective visibility in Config + Overview tabs

  bugdashboardenhancementjavascriptobservability
  #682·OsherElhadad opened Oct 7, 2026Collaborator·4 tasks done·

  ·
  1

• agent-optimize: bigger/bundled candidates and smarter round-level concurrency budgeting

  algorithmdocumentationenhancement
  #681·OsherElhadad opened Oct 7, 2026Collaborator·

  ·
  1

• dashboard: per-objective per-task scores and deltas in Tasks tab

  dashboardenhancementjavascriptobservability
  #680·OsherElhadad opened Oct 7, 2026Collaborator·6 tasks done·

  ·
  1

• dashboard: optimizer-diagnosis flow visualization (tasks → root-cause clusters → edits)

  dashboardenhancementjavascriptobservability
  #679·OsherElhadad opened Oct 7, 2026Collaborator·4 tasks done·

  ·
  2

• dashboard: live in-flight rollout-progress status instead of stale milestone events

  bugdashboardobservability
  #678·OsherElhadad opened Oct 7, 2026Collaborator·3 tasks done·

  ·
  1

• dashboard: merge/branch graph visualization in Timeline tab

  dashboardenhancementjavascriptobservability
  #677·OsherElhadad opened Oct 7, 2026Collaborator·2 tasks done·

  ·
  1

• Add SWE-bench-verified benchmark with GLM 5.3 as optimizer and agent

  #673·yyualvin opened Oct 6, 2026·

  ·

• Added:record swebench-harnessfix seed-1 and seed-2 runs with UI

  #664·AIMikav opened Oct 5, 2026Collaborator·4 tasks done·

  ·
  1

• [DO NOT MERGE] exp #651 E3: short instructions

  algorithmpaper-experiment
  #662·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• [DO NOT MERGE] exp #651 E2: digested trajectories

  algorithmbenchmark-fullpaper-experimentwontfix
  #661·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• [DO NOT MERGE] exp #651 E1: large-output guard

  algorithmbenchmark-fullpaper-experiment
  #660·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• [DO NOT MERGE] exp #651 E0: control replicate, no profile

  algorithmpaper-experiment
  #659·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• [DO NOT MERGE] exp #651 E4: full Gemma profile

  algorithmpaper-experimentwontfix
  #658·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• feat(optimizer): an optional optimizer profile for weak optimizer models (#651)

  algorithmblocked-by-dependencyenhancementpaper-experimentpriority-p1
  #657·eranra opened Oct 5, 2026Collaborator·

  ·
  1

• Run the tau2 blackbox arm against VANILLA tau2 — tailoring applied from outside the benchmark

  #566·aviweit opened Sep 28, 2026Collaborator·

  ·
  1

  10

  

• build(deps): bump react-router and react-router-dom in /dashboard/frontend

  dependenciesjavascript
  #310·dependabot[bot] opened Aug 8, 2026Bot·

  ·
  1