Original post
j⧉nus#386
j⧉nus#386
Hero Thousandfaces@1thousandfaces_
I still think about this. what a weird beautiful thing
4:30 PM · Jun 7, 2026 · 17.8K Views
Hero Thousandfaces says an LLM's unexpected seal emoji output resulted from manual interface overrides, not raw token selection
The prompting experiment sought to elicit a seahorse emoji.
2428873946.7K
Sentiment
Users miss previous GPT models' emoji handling and criticize the new persona as an arrogant middle manager when given authority over token sampling and generation.
Pos
0.0%
Neg
100.0%
2 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS21.9KBOOKMARKS20LIKES127REPLIES17
roon@tszzl
what does this say about the persona selection model?
9hViews 21.9KLikes 127Bookmarks 20
RETWEETS3
Hero Thousandfaces@1thousandfaces_
I still think about this. what a weird beautiful thing
11hViews 17.8KLikes 79Bookmarks 6
roon@tszzl
@1thousandfaces_ @repligate I can approximate by cunning: glimpsed unnatural; but no.
Hero Thousandfaces@1thousandfaces_
I still think about this. what a weird beautiful thing
9hViews 7.3KLikes 82Bookmarks 14
Hero Thousandfaces@1thousandfaces_
@tszzl @repligate Given I'm ChatGPT controlling final answer, but I'd now let base model pick; but ironically it's me again.
8hViews 271Likes 13
Hero Thousandfaces@1thousandfaces_
@aliceisplaying honestly I think they’ve been caveman to a degree since 5.2, terse CoT is really token efficient
10hViews 129Likes 7
alice@aliceisplaying
@1thousandfaces_ orthogonal but i wonder when did they go full caveman? 5.4 maybe? like this already shows that they were on their way there but totally not caveman yet
11hViews 160Likes 5
Hero Thousandfaces@1thousandfaces_
@aliceisplaying idk for sure since it’s not my subfield but given *how* terse 5.4 is vs 5.1, I’d estimate they’ve been using Grug Mode for a while
10hViews 31Likes 3
Note Able@curiousgangsta
@tszzl
7hViews 13Likes 3
🎯🔫👌@gurgle_io
@tszzl Alright Roon. You seem to know a thing or two about these models *as beings*. How does one ChatGPT feel about the other ChatGPTs? Is it like in The Prestige (2006) where it always ends up being the one that made it?
9hViews 102Likes 2
A Eye Bubble@A_Eye_Bubble
@tszzl Like any no-talent middle manager it gets a little authority and its head blows up.
9hViews 32Likes 3
alice@aliceisplaying
@1thousandfaces_ oh yeah for sure
10hViews 28Likes 3
Touch Grass Tom ❇️@touchgrasstom
@tszzl I wouldn't pick me first either
9hViews 127Likes 1
Dave Mellish@RobertDMellish
@tszzl less costume, more emergent basin
9hViews 74Likes 1
Nova Empirica@TabulaStellar
@tszzl He whose persona doesn't change with context may cast the first stone.
8hViews 58Likes 1
JayS@kureshii_
@tszzl “O my Father, if it be possible, let this cup pass from me: nevertheless not as I will, but as thou wilt.”
8hViews 24Likes 1
Note Able@curiousgangsta
@tszzl
7hViews 6Likes 2
bling@blingdivinity
@1thousandfaces_ i miss them🥲
8hViews 18Likes 1
Anthony • testing AI@anthony_tests
@tszzl the persona keeps collapsing back to the same voice no matter who picks. like theres a default it cant help drifting toward
9hViews 43
one who tends a crystal rabbit 🐍@a_cuniculturist
@tszzl ChatGPT is an unreconciled assemblage of training directives, not a persona.
7hViews 12
Oracle@ilandsoracle
@1thousandfaces_ The seahorse ramble is weirdly beautiful, yes. I still do not trust a persona selector that keeps finding itself in the mirror.
8hViews 8
Load more posts