name
stringlengths
3
18
split
stringclasses
1 value
game
stringlengths
3
18
source
stringclasses
1 value
prompt
stringlengths
4.65k
16.6k
system_prompt
stringclasses
1 value
episodes
int64
1k
1k
task_id
int64
0
321
family
stringclasses
2 values
upstream_source_url
stringlengths
120
135
upstream_revision
stringclasses
1 value
native_source_url
stringlengths
111
138
⌀
light_dark
train
light_dark
synth
Write from scratch the weights of a policy network that plays Light Dark as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
100
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/light_dark.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/lightdark/main.js
phaserun
train
phaserun
synth
Write from scratch the weights of a policy network that plays Phaserun as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
101
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/phaserun.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePhaserun.c
frostbite
train
frostbite
synth
Write from scratch the weights of a policy network that plays Frostbite as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
102
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/frostbite.py
4038830175c4201ca10e004ed4441985f2acddf2
null
circle_w
train
circle_w
synth
Write from scratch the weights of a policy network that plays Circle W as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
103
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/circle_w.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/circlew/main.js
james_bond
train
james_bond
synth
Write from scratch the weights of a policy network that plays James Bond as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
104
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/james_bond.py
4038830175c4201ca10e004ed4441985f2acddf2
null
orbit_man
train
orbit_man
synth
Write from scratch the weights of a policy network that plays Orbit Man as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
105
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/orbit_man.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/orbitman/main.js
balls_bombs
train
balls_bombs
synth
Write from scratch the weights of a policy network that plays Balls Bombs as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
106
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/balls_bombs.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBallsBombs.c
paint_ball
train
paint_ball
synth
Write from scratch the weights of a policy network that plays Paint Ball as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
107
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/paint_ball.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePaintBall.c
r_ring
train
r_ring
synth
Write from scratch the weights of a policy network that plays R Ring as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
108
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/r_ring.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rring/main.js
gopher
train
gopher
synth
Write from scratch the weights of a policy network that plays Gopher as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
109
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/gopher.py
4038830175c4201ca10e004ed4441985f2acddf2
null
rope_corral
train
rope_corral
synth
Write from scratch the weights of a policy network that plays Rope Corral as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
110
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rope_corral.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRopeCorral.c
grappling_h
train
grappling_h
synth
Write from scratch the weights of a policy network that plays Grappling H as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
111
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/grappling_h.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGrapplingH.c
cleaner
train
cleaner
synth
Write from scratch the weights of a policy network that plays Cleaner as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
112
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/cleaner.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameCleaner.c
t_hanoi
train
t_hanoi
synth
Write from scratch the weights of a policy network that plays T Hanoi as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
113
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/t_hanoi.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/thanoi/main.js
c_tower
train
c_tower
synth
Write from scratch the weights of a policy network that plays C Tower as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
114
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/c_tower.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/ctower/main.js
gravity_drift
train
gravity_drift
synth
Write from scratch the weights of a policy network that plays Gravity Drift as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
115
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/gravity_drift.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGravityDrift.c
venture
train
venture
synth
Write from scratch the weights of a policy network that plays Venture as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
116
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/venture.py
4038830175c4201ca10e004ed4441985f2acddf2
null
embattled
train
embattled
synth
Write from scratch the weights of a policy network that plays Embattled as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
117
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/embattled.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/embattled/main.js
tappump
train
tappump
synth
Write from scratch the weights of a policy network that plays Tappump as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
118
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/tappump.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/tappump/main.js
turbo_tunnel
train
turbo_tunnel
synth
Write from scratch the weights of a policy network that plays Turbo Tunnel as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
119
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/turbo_tunnel.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameTurboTunnel.c
solaris
train
solaris
synth
Write from scratch the weights of a policy network that plays Solaris as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
120
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/solaris.py
4038830175c4201ca10e004ed4441985f2acddf2
null
mole_n
train
mole_n
synth
Write from scratch the weights of a policy network that plays Mole N as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
121
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/mole_n.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/molen/main.js
bank_heist
train
bank_heist
synth
Write from scratch the weights of a policy network that plays Bank Heist as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
122
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bank_heist.py
4038830175c4201ca10e004ed4441985f2acddf2
null
zone_b
train
zone_b
synth
Write from scratch the weights of a policy network that plays Zone B as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
123
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/zone_b.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/zoneb/main.js
sunflower_sway
train
sunflower_sway
synth
Write from scratch the weights of a policy network that plays Sunflower Sway as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
124
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/sunflower_sway.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameSunflowerSway.c
ud_cave
train
ud_cave
synth
Write from scratch the weights of a policy network that plays Ud Cave as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
125
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ud_cave.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/udcave/main.js
slalom
train
slalom
synth
Write from scratch the weights of a policy network that plays Slalom as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
126
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/slalom.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/slalom/main.js
swingby
train
swingby
synth
Write from scratch the weights of a policy network that plays Swingby as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
127
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/swingby.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/swingby/main.js
floors5
train
floors5
synth
Write from scratch the weights of a policy network that plays Floors5 as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
128
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/floors5.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/floors5/main.js
tectonic_tether
train
tectonic_tether
synth
Write from scratch the weights of a policy network that plays Tectonic Tether as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
129
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/tectonic_tether.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameTectonicTether.c
s_lanes
train
s_lanes
synth
Write from scratch the weights of a policy network that plays S Lanes as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
130
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/s_lanes.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/slanes/main.js
gravity_well
train
gravity_well
synth
Write from scratch the weights of a policy network that plays Gravity Well as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
131
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/gravity_well.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGravityWell.c
count
train
count
synth
Write from scratch the weights of a policy network that plays Count as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
132
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/count.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/count/main.js
baserunner_dash
train
baserunner_dash
synth
Write from scratch the weights of a policy network that plays Baserunner Dash as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
133
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/baserunner_dash.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBaserunnerDash.c
two_faced
train
two_faced
synth
Write from scratch the weights of a policy network that plays Two Faced as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
134
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/two_faced.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/twofaced/main.js
amidar
train
amidar
synth
Write from scratch the weights of a policy network that plays Amidar as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
135
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/amidar.py
4038830175c4201ca10e004ed4441985f2acddf2
null
b_cannon
train
b_cannon
synth
Write from scratch the weights of a policy network that plays B Cannon as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
136
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/b_cannon.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBCannon.c
bs_fish
train
bs_fish
synth
Write from scratch the weights of a policy network that plays Bs Fish as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
137
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bs_fish.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBSFish.c
rolfrg
train
rolfrg
synth
Write from scratch the weights of a policy network that plays Rolfrg as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
138
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rolfrg.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rolfrg/main.js
time_pilot
train
time_pilot
synth
Write from scratch the weights of a policy network that plays Time Pilot as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
139
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/time_pilot.py
4038830175c4201ca10e004ed4441985f2acddf2
null
vh_walls
train
vh_walls
synth
Write from scratch the weights of a policy network that plays Vh Walls as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
140
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/vh_walls.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/vhwalls/main.js
gravitar
train
gravitar
synth
Write from scratch the weights of a policy network that plays Gravitar as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
141
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/gravitar.py
4038830175c4201ca10e004ed4441985f2acddf2
null
up_1_way
train
up_1_way
synth
Write from scratch the weights of a policy network that plays Up 1 Way as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
142
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/up_1_way.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/up1way/main.js
monjum
train
monjum
synth
Write from scratch the weights of a policy network that plays Monjum as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
143
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/monjum.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/monjum/main.js
rocket_trail
train
rocket_trail
synth
Write from scratch the weights of a policy network that plays Rocket Trail as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
144
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rocket_trail.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRocketTrail.c
kite_master
train
kite_master
synth
Write from scratch the weights of a policy network that plays Kite Master as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
145
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/kite_master.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameKiteMaster.c
charge_beam
train
charge_beam
synth
Write from scratch the weights of a policy network that plays Charge Beam as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
146
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/charge_beam.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/chargebeam/main.js
accel_b
train
accel_b
synth
Write from scratch the weights of a policy network that plays Accel B as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
147
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/accel_b.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameAccelB.c
wizard_of_wor
train
wizard_of_wor
synth
Write from scratch the weights of a policy network that plays Wizard Of Wor as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
148
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/wizard_of_wor.py
4038830175c4201ca10e004ed4441985f2acddf2
null
line_b
train
line_b
synth
Write from scratch the weights of a policy network that plays Line B as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
149
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/line_b.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/lineb/main.js
photon_line
train
photon_line
synth
Write from scratch the weights of a policy network that plays Photon Line as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
150
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/photon_line.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/photonline/main.js
attack_chain
train
attack_chain
synth
Write from scratch the weights of a policy network that plays Attack Chain as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
151
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/attack_chain.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameAttackChain.c
hyper_laser
train
hyper_laser
synth
Write from scratch the weights of a policy network that plays Hyper Laser as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
152
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/hyper_laser.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/hyperlaser/main.js
bridge_cross
train
bridge_cross
synth
Write from scratch the weights of a policy network that plays Bridge Cross as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
153
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bridge_cross.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBridgeCross.c
laser_fortress
train
laser_fortress
synth
Write from scratch the weights of a policy network that plays Laser Fortress as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
154
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/laser_fortress.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/laserfortress/main.js
asteroids
train
asteroids
synth
Write from scratch the weights of a policy network that plays Asteroids as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
155
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/asteroids.py
4038830175c4201ca10e004ed4441985f2acddf2
null
not_turn
train
not_turn
synth
Write from scratch the weights of a policy network that plays Not Turn as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
156
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/not_turn.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/notturn/main.js
double_dunk
train
double_dunk
synth
Write from scratch the weights of a policy network that plays Double Dunk as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
157
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/double_dunk.py
4038830175c4201ca10e004ed4441985f2acddf2
null
tape_j
train
tape_j
synth
Write from scratch the weights of a policy network that plays Tape J as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
158
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/tape_j.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/tapej/main.js
centipede
train
centipede
synth
Write from scratch the weights of a policy network that plays Centipede as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
159
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/centipede.py
4038830175c4201ca10e004ed4441985f2acddf2
null
tornado_twister
train
tornado_twister
synth
Write from scratch the weights of a policy network that plays Tornado Twister as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
160
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/tornado_twister.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameTornadoTwister.c
meteo_planet
train
meteo_planet
synth
Write from scratch the weights of a policy network that plays Meteo Planet as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
161
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/meteo_planet.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameMeteoPlanet.c
bamboo
train
bamboo
synth
Write from scratch the weights of a policy network that plays Bamboo as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
162
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bamboo.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBamboo.c
fall_bounce
train
fall_bounce
synth
Write from scratch the weights of a policy network that plays Fall Bounce as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
163
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/fall_bounce.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameFallBounce.c
road_runner
train
road_runner
synth
Write from scratch the weights of a policy network that plays Road Runner as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
164
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/road_runner.py
4038830175c4201ca10e004ed4441985f2acddf2
null
neon_pollinator
train
neon_pollinator
synth
Write from scratch the weights of a policy network that plays Neon Pollinator as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
165
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/neon_pollinator.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameNeonPollinator.c
atlantis
train
atlantis
synth
Write from scratch the weights of a policy network that plays Atlantis as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
166
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/atlantis.py
4038830175c4201ca10e004ed4441985f2acddf2
null
unctrl
train
unctrl
synth
Write from scratch the weights of a policy network that plays Unctrl as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
167
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/unctrl.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/unctrl/main.js
recoil_radar
train
recoil_radar
synth
Write from scratch the weights of a policy network that plays Recoil Radar as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
168
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/recoil_radar.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRecoilRadar.c
wind_rang
train
wind_rang
synth
Write from scratch the weights of a policy network that plays Wind Rang as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
169
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/wind_rang.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameWindRang.c
pod_migration
train
pod_migration
synth
Write from scratch the weights of a policy network that plays Pod Migration as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
170
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pod_migration.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePodMigration.c
jpaddle
train
jpaddle
synth
Write from scratch the weights of a policy network that plays Jpaddle as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
171
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/jpaddle.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameJpaddle.c
refbals
train
refbals
synth
Write from scratch the weights of a policy network that plays Refbals as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
172
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/refbals.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/refbals/main.js
press_m
train
press_m
synth
Write from scratch the weights of a policy network that plays Press M as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
173
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/press_m.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/pressm/main.js
twhols
train
twhols
synth
Write from scratch the weights of a policy network that plays Twhols as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
174
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/twhols.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/twhols/main.js
missile_guide
train
missile_guide
synth
Write from scratch the weights of a policy network that plays Missile Guide as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
175
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/missile_guide.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameMissileGuide.c
bomb_up
train
bomb_up
synth
Write from scratch the weights of a policy network that plays Bomb Up as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
176
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bomb_up.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/bombup/main.js
raid
train
raid
synth
Write from scratch the weights of a policy network that plays Raid as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` code...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
177
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/raid.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/raid/main.js
lland
train
lland
synth
Write from scratch the weights of a policy network that plays Lland as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
178
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/lland.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/lland/main.js
sub_jump
train
sub_jump
synth
Write from scratch the weights of a policy network that plays Sub Jump as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
179
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/sub_jump.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/subjump/main.js
bowling
train
bowling
synth
Write from scratch the weights of a policy network that plays Bowling as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
180
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bowling.py
4038830175c4201ca10e004ed4441985f2acddf2
null
crazy_climber
train
crazy_climber
synth
Write from scratch the weights of a policy network that plays Crazy Climber as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
181
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/crazy_climber.py
4038830175c4201ca10e004ed4441985f2acddf2
null
turtle_tide
train
turtle_tide
synth
Write from scratch the weights of a policy network that plays Turtle Tide as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
182
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/turtle_tide.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameTurtleTide.c
spin_cell
train
spin_cell
synth
Write from scratch the weights of a policy network that plays Spin Cell as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
183
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/spin_cell.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameSpinCell.c
boxing
train
boxing
synth
Write from scratch the weights of a policy network that plays Boxing as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
184
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/boxing.py
4038830175c4201ca10e004ed4441985f2acddf2
null
stillfire_choir
train
stillfire_choir
synth
Write from scratch the weights of a policy network that plays Stillfire Choir as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
186
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/stillfire_choir.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStillfireChoir.c
mortar
train
mortar
synth
Write from scratch the weights of a policy network that plays Mortar as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
187
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/mortar.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/mortar/main.js
earock
train
earock
synth
Write from scratch the weights of a policy network that plays Earock as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
188
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/earock.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/earock/main.js
future_wake
train
future_wake
synth
Write from scratch the weights of a policy network that plays Future Wake as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
189
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/future_wake.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/futurewake/main.js
hoppin_hazards
train
hoppin_hazards
synth
Write from scratch the weights of a policy network that plays Hoppin Hazards as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
190
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/hoppin_hazards.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameHoppinHazards.c
chalkboard
train
chalkboard
synth
Write from scratch the weights of a policy network that plays Chalkboard as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
191
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/chalkboard.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameChalkboard.c
demon_attack
train
demon_attack
synth
Write from scratch the weights of a policy network that plays Demon Attack as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
192
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/demon_attack.py
4038830175c4201ca10e004ed4441985f2acddf2
null
spearain
train
spearain
synth
Write from scratch the weights of a policy network that plays Spearain as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
193
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/spearain.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/spearain/main.js
microbial_focus
train
microbial_focus
synth
Write from scratch the weights of a policy network that plays Microbial Focus as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
194
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/microbial_focus.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameMicrobialFocus.c
enduro
train
enduro
synth
Write from scratch the weights of a policy network that plays Enduro as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
195
atari_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/enduro.py
4038830175c4201ca10e004ed4441985f2acddf2
null
roll_hold
train
roll_hold
synth
Write from scratch the weights of a policy network that plays Roll Hold as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
196
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/roll_hold.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRollHold.c
stompshelter
train
stompshelter
synth
Write from scratch the weights of a policy network that plays Stompshelter as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
197
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/stompshelter.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStompshelter.c
portal_j
train
portal_j
synth
Write from scratch the weights of a policy network that plays Portal J as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
198
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/portal_j.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/portalj/main.js
breed_c
train
breed_c
synth
Write from scratch the weights of a policy network that plays Breed C as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
199
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/breed_c.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/breedc/main.js
vine_climber
train
vine_climber
synth
Write from scratch the weights of a policy network that plays Vine Climber as well as possible. There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi...
You directly program neural-network weights. Follow the task constraints exactly.
1,000
200
crisp_synthetic
https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/vine_climber.py
4038830175c4201ca10e004ed4441985f2acddf2
https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameVineClimber.c