GPT-4.5 or GPT-5 being tested on LMSYS?

OpenAI could either hire private testers or use AB testing on ChatGPT Plus users (for example, oftentimes, when using ChatGPT, I have to select between 2 different responses to continue a conversation); both are probably much more better (in many aspects: not leaking GPT-4.5/5 generations (or the existence of a GPT-4.5/5) to the public at scale and avoiding bias* (because people probably rate GPT-4 generations better if they are told (either explicitly or implicitly (eg. socially)) it's from GPT-5) to say the least) than putting a model called 'GPT2' onto lmsys.

* While lmsys does hide the names of models until a person decides which model generated the best text, people can still figure out what language model generated a piece of text** (or have a good guess) without explicit knowledge, especially if that model is hyped up online as 'GPT-5;' even a subconscious "this text sounds like what I have seen 'GPT2-chatbot' generate online" may influence results inadvertently.

** ... though I will note that I just got a generation from 'gpt2-chatbot' that I thought was from Claude 3 (haiku/sonnet), and its competitor was LLaMa-3-70b (I thought it was 8b or Mixtral). I am obviously not good at LLM authorship attribution.

In my runs even Llama3-8B-chat gets it right.

A dolphin/mistral fine tune also go it right.

Deepseek 67B also.

It strikes me as unprecedented that a technology which takes arbitrary language-based commands can actually surface and synthesize useful information, and it gets better at doing it (even according to extensive impartial benchmarking) at a fairly rapid pace. It’s technology we haven’t really seen before recently, improving quite quickly. It’s also being adopted very rapidly.

I’m not saying it’s certainly the fastest growth of all time, but I think there’s a decent case for it being a contender. If we see this growth proceeding at a similar rate for years, it seems like it would be a clear winner.

Wow.

I did the arena and kept asking this:

Has Anyone Really Been Far Even as Decided to Use Even Go Want to do Look More Like?

All of them thought it was gibberish except gpt2-chatbot. It said:

The phrase you're asking about, "Has anyone really been far even as decided to use even go want to do look more like?" is a famous example of internet gibberish that became a meme. It originated from a post on the 4chan board /v/ (video games) around 2009. The post puzzled readers due to its nonsensical structure, sparking amusement and confusion across the internet community.

*Interpretation and Usage:*

1. *As a Meme:* This phrase is typically used online to mock confusing, poorly worded, or illogical statements. It’s often brought up in forums or comment sections as a humorous way to question the coherence of someone else’s words or to deliberately confuse someone further.

2. *Cultural Impact:* It reflects a broader internet culture that values absurdity and nonsensical humor. Memes like this are a way for internet communities to bond over shared "inside jokes" or a common understanding of internet history and culture.

3. *In Communication:* You might see it used in response to something that doesn't make sense, or when someone wants to playfully indicate that what they're saying or asking isn't meant to be taken seriously. It’s a playful embrace of chaos and confusion that’s typical in many online communities.

*Use in Conversation:* If you're in a context where internet memes and humor are appreciated, dropping this phrase could serve as a light-hearted way to engage or nudge others into a fun, if a bit bewildering, exchange. Just be aware that its usage is best received among those familiar with meme culture and internet history.

You asked it to pretend.