tweetindex
ES

James Whittington ✓

@jcrwhittington

Neuroscience, AI, physics, medicine.

2 693Followers
184Following
445Posts total
1.2MViews on collected posts

Últimas publicaciones

@fchollet @arcprize 2.2K views · 11 likes · 0 reposts · 0 replies Open on X →
@fchollet what do you think? Our attempt at text-based arc-agi-3 set of games :) 2.5K views · 24 likes · 0 reposts · 2 replies Open on X →
@jcrwhittington Nice! "Each is beaten by at least one human on their first attempt" — what pass at K is that (i.e., how many humans attempted)? And what's the population you sourced from? 1.8K views · 13 likes · 0 reposts · 1 replies Open on X →
The 21 public games are playable in your browser, and the API lets you run any model against them. No pressure, but we know humans can solve all of these on their first attempt :) https://t.co/EsdewBxS6W If you enjoy the puzzles, share them with a friend, or see if you can 3.8K views · 36 likes · 3 reposts · 3 replies Open on X →
We find that 1) Opus 5 & Fable marked a real increase in ability on these games, but they still struggle with the hardest tiers of games (Tiers 6 and 7) 2) Agentic harnesses like Prime don’t offer any benefit over a basic (non coding!) harness 3) There’s a large gap between https
14.8K views · 55 likes · 5 reposts · 3 replies Open on X →
Each of the 70 games is a self-contained, and different, world rendered as a short string, e.g., like this where you can “walk” around as the arrow 1____A___→_____2 () The rules and win conditions are hidden, you have to discover it all by experimentation. The games are https:
4.8K views · 37 likes · 4 reposts · 2 replies Open on X →
For AI to chart new science and genuinely make new discoveries, more than just combining known tools together, it first needs to prove it can make discoveries in the easiest possible setting. So, we present, DiG-bench, a benchmark of 70 discovery games, each based in text and ht
7.1K views · 50 likes · 6 reposts · 3 replies Open on X →
We’re excited to announce DiG-bench, a new benchmark for discovery! Over the last few weeks we’ve been testing frontier AI models on our novel discovery games and seeing how they score. Each game is a text-based environment, so they probe discovery capabilities in the natural h
1.2M views · 1.2K likes · 360 reposts · 44 replies Open on X →

Frente a cuentas del mismo tamaño

8 publicaciones de los últimos 90 días, junto al rango de under 10K seguidores. por debajo de sus pares en alcance y en interacción.

Visualizaciones medianas4 276esta cuenta4 175mediana de under 10K
Alcance, %158.80%esta cuenta250.82%mediana de under 10K
Interacción, %0.81%esta cuenta1.41%mediana de under 10K
MétricaEsta cuentaMediana de under 10KProporción
Visualizaciones medianas por publicación4 2764 1751.02×
Alcance (visualizaciones ÷ seguidores)158.80%2.5× audience0.63×
Tasa de interacción0.81%1.41%0.57×

Otras cuentas de este rango →   Comparar con otra cuenta →   Cómo se construyen estas referencias →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

1.2M12 Aug
7.1K
4.8K
14.8K
3.8K
1.8K
2.5K
2.2K

Last 8 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.15%12 Aug
0.84%
0.90%
0.45%
1.15%
0.77%
1.05%
0.50%

Reactions — likes, reposts, replies and quotes — divided by views. Median for under 10K accounts is 1.41%.

What the audience does

Likes56.7%1 418 in total
Reposts15.1%378 in total
Replies2.3%58 in total
Quotes10.0%250 in total
Bookmarks15.9%397 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Cuentas similares