HeliumTrades/helium-model-worldview-benchmark
Viewer • Updated • 323 • 629 • 1
Helium Trades
HeliumTrades
AI & ML interests
None yet
Recent Activity
new activity 1 day ago
BGPT-OFFICIAL/refute:Human scientific value note (why this is not another skill board) new activity 1 day ago
OpenEvals/README:New Benchmark DatasetOrganizations
None yet
Human scientific value note (why this is not another skill board)
#28 opened 1 day ago
by HeliumTrades
New Benchmark Dataset
🚀 5
20
#2 opened 6 months ago
by 
burtenshaw
Why Market Resolution sits next to Model Worldview
#9 opened 1 day ago
by HeliumTrades
liked a
dataset 1 day ago
Why REFUTE exists: scientific-sounding ≠scientifically honest
1
#26 opened 1 day ago
by HeliumTrades
Related open dataset: values, refusal, cue sensitivity
#2 opened 1 day ago
by HeliumTrades
What we found on values, politics, and refusal
#9 opened 1 day ago
by HeliumTrades
Show the labeled efficient frontier
1
#16 opened 1 day ago
by HeliumTrades
Lead with Truth Score and the efficient frontier
1
#25 opened 1 day ago
by HeliumTrades
updated 2
datasets 1 day ago
Prefer live dataset scores; refresh intro and assets
1
#15 opened 3 days ago
by HeliumTrades
Polish: sync stale docs, drop obsolete four_pillars graphic
1
#24 opened 3 days ago
by HeliumTrades
Refresh assets after ITT rescoring and full 15-model board
1
#14 opened 3 days ago
by HeliumTrades
Final pass: ITT scoring, full board graphics, bottom-three story
1
#23 opened 3 days ago
by HeliumTrades
Refresh scatter after corrected Gemma scores
1
#13 opened 3 days ago
by HeliumTrades
Fix Gemma truncation bug and four-questions graphic
1
#22 opened 3 days ago
by HeliumTrades
Use the live dataset as the single score source
1
#12 opened 3 days ago
by HeliumTrades
Restart leaderboard against the completed dataset
1
#11 opened 3 days ago
by HeliumTrades