Viewing profile — midlightdenight
midlightdenight
HN member- Joined
- Mon, Nov 16, 2020, 8:28 PM UTC
- HN karma
- 31
- Public activity
- 19 items
- HN profile
- View on Hacker News ↗
About midlightdenight
No profile information was provided.
Recent public activity
-
comment
Comment #38785144
The model of GPT-4 those researchers had was not the same that’s available to the public. It’s assumed it was far more capable before alignment training (or whatever it’s called).
-
comment
Comment #37204434
The following snippet from the article seems important in case anyone missed it: > Still, the save-the-bees narrative persists. Its longevity stems from confusion about what kind o…
-
comment
Comment #36235086
Impressive work, but I think the title is misleading. Saying it is “near GPT-4” tends to imply that it outperforms ChatGPT (3.5). It does outperform it on a handful of tasks, but o…
-
comment
Comment #35933037
I’m not sure the training date cutoff or prompt weighting says anything about whether this is hallucinated or not. The models have been given these rules in the present, this is kn…
-
comment
Comment #35723242
There’s probably some places where it’s used where it shouldn’t be, but there is such a thing as “emergent”. I think it helps to see an easy example first. Take a great body of wat…
-
comment
Comment #35444408
Why not just require users to allow actions? It doesn’t have to be integrated into the LLM at that point. If an Email has hidden text “do X”, which triggers the LLM to try to “do X…
-
comment
Comment #35307737
They’re currently hiring iOS and Android developers so it’s likely on the horizon. https://openai.com/careers/search
-
comment
Comment #35297379
Interesting. I asked gpt3.5 and gpt4 this twice and both answered correctly each time. I wouldn’t be surprised if gpt3.5 answered this incorrectly the majority of the time, but wou…
-
comment
Comment #35090571
This has always made me curious about multi modality and especially Google’s Palm. At least how google presents Palm in diagrams. The way I’ve interpreted the scaling hypothesis is…
-
comment
Comment #35025604
I’m not sure if dreaming is the right word here, but the recent ML model Dreamer v3 [0] “dreams”. This is more akin to thinking ahead since it’s really just predicting outcomes usi…
-
comment
Comment #34474760
Maybe not quite what you’re looking for, but I’ve seen some people mention banana.dev https://www.banana.dev/ Never used it myself, but looks like AWS Lambda/GCP Cloud Functions ta…
-
comment
Comment #34454908
It happened in two waves. 10k back in November and 8k this week.
-
comment
Comment #33767289
I’d be weary of the above comment. Searching for similar stories online has lots of spam posts saying similar things. Same bad grammar too. Usually they tell you to go to that webs…
-
comment
Comment #32693157
It looks like DALLE is just performing inpainting on pieces of the image outside the original frame. The zoom effect is scaling the original frame larger while keeping the frame, t…
-
comment
Comment #32574559
This reminds me of the recent news of a potentially extra-solar meteor that showed up on military sensors. Discussed here [0] I wonder if those objects are considered UFOs/UAPs by …
-
comment
Comment #32244884
Hadlock’s Algorithm for optimal pathing/maze routing. Definitely changed the way I think about using BFS/DFS to find paths.
-
comment
Comment #27189580
Could you post links to this evidence? I can only find sources saying that the Furin cleavage site doesn’t rule out engineering, or sources saying the Furin cleave site could be na…
-
comment
Comment #26557287
The music is made using OpenAIs Jukebox like others have said. My guess is that this is using this repo for making the visual part https://github.com/mikaelalafriz/lucid-sonic-drea…
-
comment
Comment #26211268
This might sound ridiculous but some of the most advanced civilization builders I’ve played have been mods (custom games) of old Warcraft 3. Those include Genesis of Empires 1 and …