lforla
PulseAugur coverage of lforla — every cluster mentioning lforla across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
DeepSeek-V4.1-Flash-1017 launches with 1M context, Haiku 5.5 cost guard, and LFORLA evaluation details
Several AI-related developments are emerging, including the DeepSeek-V4.1-Flash-1017 model, which offers a 1 million token context window and an MIT license, showing significant improvements on the Terminal-Bench benchm…
-
New LFORLA benchmark tests LLMs on election forecasting accuracy
A new benchmark called LFORLA has been developed to evaluate Large Language Models (LLMs) on their ability to forecast election outcomes. This benchmark requires models to commit to specific predictions for future elect…
-
HY3 model edges out Nemotron 3 Ultra in bias and stereotype audit
A recent audit by lforla using the Bias Stereotypes benchmark revealed that the HY3 model outperformed Nemotron 3 Ultra by a narrow margin. HY3 achieved a score of 82.9 compared to Nemotron's 81.7, primarily due to supe…
-
Nemotron 3 Ultra leads new AI recruitment benchmark with 90.87 score
The new Team Recruitment (Oracle) benchmark, developed by lforla, evaluates AI agents on their ability to query a résumé oracle and assemble a team that meets specific criteria. Currently, the leaderboard features two f…