tweetindex
DE

Edward Grefenstette

@egrefen · London, United Kingdom · joined 29 Apr 2007

FR/US/GB AI/ML Person, Director of Research at @GoogleDeepMind, Honorary Professor at @BOLD_Lab_AI, @ELLISforEurope Fellow. All posts are personal.

46 475Followers
926Following
16 152Posts total
290.2KViews on collected posts

Neueste Beiträge

Disappointed to hear the @TfL announcement system in King's Cross blast an @audible_com announcement formatted as a normal announcement. Misleading, tasteless, and unwelcome. 3.1K views · 18 likes · 0 reposts · 0 replies Open on X →
"SPCX to the moon 🚀🚀🚀" bros have really gone silent, haven't they 😅 3.7K views · 5 likes · 1 reposts · 3 replies Open on X →
So... solving reward hacking necessary for RSI, you say? 🤔😉 5.1K views · 19 likes · 0 reposts · 2 replies Open on X →
It's recursive self-improvement if it works, reward hacking if it doesn't. 18.8K views · 208 likes · 9 reposts · 2 replies Open on X →
Amazing launch, @redpony and team! Great to see Google ship technology in this space 🥰 3.6K views · 17 likes · 1 reposts · 2 replies Open on X →
@egrefen Because I consider you a researcher and scientist this concerns and troubles me deeply: "...or ever?..." 240 views · 1 likes · 0 reposts · 0 replies Open on X →
As always, I end this thread with a reference to last year's. Happy end of 2025, everyone, and best wishes for the new year! 🥳🎉 [12/12] https://t.co/vjYiV0Cx1b 5.4K views · 35 likes · 0 reposts · 7 replies Open on X →
On the personal front, I've been focussing on trying to be a good dad, become a better golfer, and be more organized. I did two years of BuJo for task management and am think about doing something more integrated into my Obsidian workflow in the new year.… [11/12] 4.3K views · 43 likes · 0 reposts · 3 replies Open on X →
We also published https://t.co/QNs4JcCnME about a year or so after the work was done. I can't comment on if, where, or how this might be used internally. [10/12] 3.9K views · 44 likes · 3 reposts · 1 replies Open on X →
Along the way we published a few technical reports such https://t.co/BTs8zoiL2Y or https://t.co/G8oalLHWrb, which can be seen as hints of the broader directions we are looking at. [9/12] 5.1K views · 49 likes · 4 reposts · 1 replies Open on X →
We've also made some progress in terms the actual learning process itself, allowing open-ended agents, at the instance level, to learn and adapt with human-like data efficiency. This potentially points at a broader way of improving agents at scale, which we are working on. [7/12] 75.1K views · 164 likes · 11 reposts · 5 replies Open on X →
On the more theoretical front, we've been interested in what it actually *means* to have complementary/cooperative agents, rather replacements of our own agency, and also how agents could self-motivate. I want to stress that these are both unsolved, open questions. [8/12] 5K views · 60 likes · 3 reposts · 1 replies Open on X →
I can't say much more, but I did give some general high level directions of interest at keynote I gave the SEA workshop at NeurIPS: https://t.co/I5n9bVQhPl Perhaps the video will be put online in the next month or so... [5/12] 5.4K views · 58 likes · 6 reposts · 2 replies Open on X →
Broadly, we've been making good progress with regard to how open-ended agents can learn "in the wild", with less human intervention in their learning process, while still ensuring they remain aligned with human behaviors and interests. [6/12] 12.4K views · 85 likes · 4 reposts · 1 replies Open on X →
With my broad circle of collaborators, we've investigated some interesting questions at the frontier of what's possible with contemporary AI, to (potentially) power the models of tomorrow and (maybe) some of today. [4/12] 4.7K views · 26 likes · 1 reposts · 2 replies Open on X →
At @GoogleDeepMind I continue to collaborate with a broad set of wonderful people across Frontier AI and Gemini teams. In particular with @SianGooding, Lucia Lopez, João Araújo, Yusra Ibrahim, @MartinKlissarov, @jingling_li, @HolarisSun, and many others I see near daily. [3/12] 5.1K views · 32 likes · 3 reposts · 2 replies Open on X →
🧵 Time for a short end-of-2025 wrap up 🧵 Genuinely aiming for this to be a short one for two reasons: 1. I'm doing it at the last moment 😅 2. Most of what I was involved in is not stuff that can be shared publicly (yet… or ever?). Or maybe I was just lazy... Let's go [1/12] 81.8K views · 228 likes · 21 reposts · 7 replies Open on X →
On the UCL front, @akbirkhan and @_robertkirk, who had started at Anthropic and AISI, formally finished their PhDs early in the year, leaving me with two PhD students. @LauraRuis completed hers more recently and is now doing a postdoc at MIT. Down to one… [2/12] 5.4K views · 40 likes · 3 reposts · 1 replies Open on X →
🧵 As 2024 wraps up, please pardon my usual self-indulgence in tweeting about the year gone by. 🧵 This will be a reasonably short one... OR WILL IT? [1/17] 42K views · 35 likes · 1 reposts · 6 replies Open on X →

Im Vergleich zu Konten gleicher Größe

5 Beiträge aus den letzten 90 Tagen, verglichen mit der Größenklasse 10K–100K Follower. wird weit gezeigt, aber nur wenige dieser Zuschauer reagieren.

Medianaufrufe3 661dieses Konto924Median für 10K–100K
Reichweite, %7.88%dieses Konto3.62%Median für 10K–100K
Interaktion, %0.56%dieses Konto1.52%Median für 10K–100K
KennzahlDieses KontoMedian für 10K–100KVerhältnis
Medianaufrufe pro Beitrag3 6619243.96×
Reichweite (Aufrufe ÷ Follower)7.88%3.62%2.18×
Interaktionsrate0.56%1.52%0.37×

Weitere Konten dieser Größe →   Mit einem anderen Konto vergleichen →   Wie diese Vergleichswerte entstehen →

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

12.4K31 Dec
5.4K
5K
75.1K
5.1K
3.9K
4.3K
5.4K
2402 Jan
3.6K12 Aug
18.8K18 Aug
5.1K19 Aug
3.7K21 Aug
3.1K3 Sep

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.75%31 Dec
1.23%
1.27%
0.26%
1.07%
1.22%
1.08%
0.77%
0.42%2 Jan
0.56%12 Aug
1.17%18 Aug
0.41%19 Aug
0.27%21 Aug
0.58%3 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes70.0%1 167 in total
Reposts4.3%71 in total
Replies2.9%48 in total
Quotes1.5%25 in total
Bookmarks21.4%357 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.

Ähnliche Konten