Select the second-largest of four inputs. Keeping the top two candidates is enough; fully sorting is another option.
In TIS-100 Bench, 16 LLMs each had one attempt at this TIS-100 puzzle, with 45 minutes to find a working program and keep improving it, without seeing the community leaderboard. 16 of 16 models solved it. Together they found 4 new Pareto-frontier points beyond the community leaderboard.
16 / 16models solved it
414best LLM cycles (336 community)
4new frontier points
Records
Score
Community best
Best LLM
Model
Cycles
336
414
Kimi K3.1 Pro· +23%
Nodes
6
6
GPT-6 AstraTies record
Instructions
31
40
GPT-6 Astra· +29%
Community records from the community leaderboard when we froze it. A record counts only after fresh-game replay and 100,000 passing random tests.
Pareto frontier
Pareto explorer
Lower is better ↙
Loading interactive Pareto explorer…
Community leaderboardModel frontier pointNew frontier point
Select a point with the mouse or keyboard to inspect its scores and program.
Search over time
Loading puzzle trajectories…
The search over time
Best in-game score so far, before independent testing. Each score can come from a different program, and lines stop when the episode ends.
MOV UP ACC
MOV ACC RIGHT
MOV ACC RIGHT
MOV ACC RIGHT
MOV ACC RIGHT
Node 1
S: MOV UP ACC
SUB LEFT
JLZ A
ADD LEFT
MOV ACC RIGHT
MOV ACC RIGHT
MOV LEFT RIGHT
MOV LEFT NIL
JMP S
A: MOV LEFT RIGHT
MOV LEFT RIGHT
ADD LEFT
MOV ACC RIGHT
Node 2
S: MOV UP ACC
SUB RIGHT
JLZ A
ADD RIGHT
MOV ACC DOWN
MOV LEFT DOWN
MOV RIGHT DOWN
JMP M
A: MOV RIGHT DOWN
MOV LEFT DOWN
ADD RIGHT
MOV ACC DOWN
M: MOV LEFT DOWN
MOV LEFT DOWN
Node 3
MOV UP ACC
MOV ACC LEFT
MOV ACC LEFT
MOV ACC LEFT
Node 6
MOV UP ACC
SAV
SUB UP
JLZ A
MOV UP ACC
MOV UP DOWN
MOV UP NIL
JMP O
A: SWP
MOV UP NIL
MOV UP NIL
MOV UP DOWN
O: MOV ACC DOWN
MOV ACC DOWN
Node 9
S: MOV UP ACC
SUB UP
JGZ A
MOV UP DOWN
JMP S
A: ADD UP
MOV ACC DOWN
MOV LEFT DOWN
MOV LEFT ACC
MOV ACC DOWN
MOV ACC DOWN
MOV -10 DOWN
MOV UP ACC
MOV ACC DOWN
MOV ACC DOWN
MOV -10 DOWN
MOV RIGHT ACC
MOV ACC DOWN
MOV ACC DOWN
MOV -11 DOWN
Node 3
MOV UP LEFT
Node 6
MOV UP ACC
L: SAV
SUB UP
JGZ A
MOV UP ACC
SWP
JMP O
A: MOV UP ACC
O: MOV ACC DOWN
MOV ACC DOWN
SWP
JRO UP
Node 9
MOV UP NIL
MOV UP ACC
SUB UP
SUB 999
ADD 999
ADD UP
SUB UP
SUB 999
ADD 999
ADD UP
MOV ACC DOWN