pokerl
AI
reinforcement learning
SHOW HN2025
183points
68comments
#368of 3,837 in 2025
Top 10%of 2025 by points
How HN took it
CuriousMostly questions and neutral comments
7 positive, 23 neutral or asking, 0 mixed, 4 critical, out of 34 comments about the project. Thread score +0.15.
Idea: 5 for, 0 againstDesign: 0 for, 2 againstPerformance: 0 for, 1 againstDocs: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Impressive small-model RL work
From 3 comments
Wow nice work. 10M is a tiny model and I suspect this might be the future for specialised work. I can also imagine the progress towards AGI/ASI to have smaller models used as…
N_Lens on HNVery nice! Nice to see demonstrations of reinforcement learning being used to solve non-trivial tasks.
jononor on HNConsidering how many things are less complicated than Pokemon, this is very cool
nimish on HN
Shows RL works without LLMs
From 1 comment
Note: What makes this interesting is that this is a pre-LLM project which shows that in some projects you don't need an "LLM" for this. All you need is just a plain old…
rvz on HN
Objections
Battling is unimportant to the agent's progress
From 2 comments
a cool idea, except that battling actually doesn't even matter to the ai. if you look at what the agent is doing during a battle, it is sort of spamming options + picking…
xinpw8 on HNas it stands, battling is wholly unimportant to completing the game, as long as the agents can eventually complete the trainer battles mandatory for plot advancement. it's funny…
xinpw8 on HN
Page navigation bug fixed after report
From 2 comments
Heads up, clicking "Next Page" just takes you to an empty screen, you have to use the navigation links on the left if you want to get read past the first screen.
worble on HNThanks for the heads up. I just pushed a fix.
drubs on HN
Requests
Stream gameplay to Twitch
From 2 comments
Please stream the gameplay to twitch so people can compare.
novia on HNCan you make a twitch stream of a single agent playing?
novia on HN
Add arxiv link
From 1 comment
Really missing the arxiv link. The whole page reads like the arxiv link should be in the next paragraph, but it never appeared.
KeplerBoy on HN
Questions
Can an LLM be bolted onto the RL agent?
From 2 comments
Really cool work. It seems like some critical areas (team rocket, safari zone) rely on encoding game knowledge into the reward function somehow, which "smuggles in" external…
levocardia on HNFor well-known games like "Pokemon Red" I wonder how much of that game knowledge would be "smuggled in" by an LLM in it's training data if you just replaced the external info in…
Xelynega on HN
How were the game and environment wrapped for RL?
From 2 comments
What an awesome project! I'm curious - I would have thought that rewarding unique coordinates would be enough to get the agent to (eventually) explore all areas, including the…
bubblyworld on HNyou dont port it you wrap it. you can put anything in an rl environment. usually emulators are done with bizhawk, and some lua. worst case theres ffi or screen capture.
wegfawefgawefg on HN
How many button presses are needed to beat the game?
From 2 comments
This is very cool, congrats! I wonder, does anyone have a sense of the approximate raw number of button presses required to beat the game? Mostly curious to see how that compares…
differintegral on HNI imagine < 10000. https://github.com/KeeyanGhoreshi/PokemonFireredSingleSequence and https://www.youtube.com/watch?v=6gjsAA_5Agk. I believe this is something like 200k and is a…
tarentel on HN
Alternatives named
Claude Plays Pokémon
From 1 comment
Claude 3.7 recently failed to finish Pokemon after getting stuck in a corner and deciding it was impossible to get out
mclau156 on HN
Pi Plays Pokémon Sapphire
From 1 comment
Judging by the “pi plays Pokemon Sapphire”, uh, not in a reasonable amount of time? It’s been at it for over 3 years, hasn’t gotten a gym badge yet, mostly stays in the starting…
drdeca on HN