November 2026
87 stories so far · page 5 of 5Thursday 12 November
17:56AgentsIIBenchmarked
A new benchmark built from real researcher sessions finds the best of seven models fully completes only two-thirds of tasks. Scores fall sharply once earlier decisions and existing files are checked.
17:56Vision & multimodalIIBenchmarked
A preprint benchmark of 12 image-counting tasks finds that free, off-the-shelf vision-language models are a strong starting point for estimating totals with a small human-labelled sample. Letting the model pick which images people check usually helps further.
17:56AgentsIIBenchmarked
A preprint proposes that AI agents should treat each lesson learned from a task as a hypothesis to test, and keep it only if it helps without breaking anything. The benchmark gains are large, but the comparisons need care and no code is released.
17:56AgentsIIBenchmarked
A preprint describes EngineAgent, a set of AI agents that adapts on-device inference engines to new language models. It passed 9 of 12 tasks against 4 for the best general coding agent. The benchmark is small and the authors built it themselves.
17:56Vision & multimodalIIIOpen source
A single-author preprint retrains a small open streaming model so its median word appears about 90 ms before the speaker finishes it, each word tagged with a usable confidence. The cost is more words taken back, and it is less accurate than larger models.
17:56RoboticsIIIOpen source
A preprint from the National University of Singapore folds a robot's actions, joint readings and value estimates into a single frame that a pretrained video model treats as natural. Queries run 39% faster and simulated benchmark scores rise slightly.
17:56EfficiencyIIIOpen source
A single-author preprint argues that popular losses for training draft models measure the wrong thing, and proposes an exact alternative. The gains on one 35B model are consistent but small.
All months