37 min
Ship the PaperEvaluation
Safe, because it looked everything up
A new paper lets an AI judge keep only the verdicts it is sure of, with a stated cap on how many can be wrong. I rebuilt it on a laptop to find where that cap stops holding.
Ship
All writing
Section
Ship the Paper
One research paper at a time: rebuild the central idea, test it, and say whether it survives production.
Essays
- 32 min
Ship the PaperEfficiency & Serving
Right company, right quarter, wrong question
FinCacheServe lets a financial assistant reuse finished answers until a filing changes. I rebuilt its reuse check to see whether an up-to-date answer is also the right one.
Wait - 31 min
Ship the PaperLong-Horizon Planning & Reliability
Branch six was never on the map
Dream-RSI scores a new exploration policy by replaying a finished discovery run instead of paying for another one. I rebuilt the replay to see what it can pick, and what it never sees.
Wait - 42 min
Ship the PaperLong-Horizon Planning & Reliability
I added 100 to an invoice and the score did not move
A new paper scores a context summary by the blocked or repeated tool calls it causes next. I rebuilt the scorer to find out how much of a summary's damage ever makes that noise.
Wait - 32 min
Ship the PaperSafety & Security
The cheapest attack never got a number
Ehghaghi and colleagues price a jailbreak by the compute it burns instead of the queries it sends. I rebuilt their two summary metrics to find out which one a nightly build could trust.
Wait - 37 min
Ship the PaperAgents
Find out who wrote down the diff
Swap the harness around a coding agent and its score moves nearly as far as swapping the model. I rebuilt the step behind most of that gap: how the agent's finished patch gets collected.
Ship - 32 min
Ship the PaperAgents
The page that never changed
FCPAgent makes a browser agent write down, before each step, what would prove that step wrong. I rebuilt the checking half and found the one failure no such statement can see.
Ship - 25 min
Ship the PaperAgents
The cockpit nobody designed on purpose
AgentOccam shrinks what a browser agent reads and what it is allowed to do. I rebuilt the shrinking step and it held up; the question it never asks is whether the action is allowed.
Wait