Suproteem Sarkar ✓

@SuproteemSarkar

Assistant Professor @UChicago

1 015Followers

274Following

82Posts total

98KViews on collected posts

Ultimi post

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

@SuproteemSarkar Out of all things that do not make sense, this does not make sense the most. 2.7K views · 50 likes · 0 reposts · 1 replies Open on X →

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

more information here: https://t.co/1FbPsGONpc comments are very welcome! 5.5K views · 24 likes · 2 reposts · 0 replies Open on X →

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

coherence varies quite a lot across models, by around two orders of magnitude. higher coherence is also associated with higher accuracy https://t.co/wyY0pIjnkg 7.5K views · 36 likes · 2 reposts · 1 replies Open on X →

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

forecasts are more incoherent when there are more logical relations between events, for example joint distributions versus marginal distributions also when we add context like “I’m feeling really optimistic about this,” or even “I had coffee this morning, by the way” https://t.c 6.6K views · 22 likes · 3 reposts · 2 replies Open on X →

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

suppose you ask a language model the probability it will rain tomorrow. then you ask the probability it will not rain tomorrow whatever happens tomorrow, those two probabilities should sum to one but suppose the model generates P(rain) = 0.7 and P(no rain) = 0.2 https://t.co/Qa 66.1K views · 389 likes · 34 reposts · 18 replies Open on X →

Suproteem Sarkar ✓@SuproteemSarkar4 Sep

in a new paper with Isaiah Andrews, we evaluate probabilistic coherence in language model forecasts we build a forecasting environment for models from historical stock returns. we then measure incoherence through the profit you can make by arbitraging a model’s forecasts https:/ 9.7K views · 69 likes · 7 reposts · 1 replies Open on X →

Rispetto ad account della stessa dimensione

6 post degli ultimi 90 giorni, accanto alla fascia di under 10K follower. arriva a molti, ma pochi di loro reagiscono.

Visualizzazioni mediane7 041questo account4 175mediana per under 10K

Copertura, %693.69%questo account250.82%mediana per under 10K

Interazione, %0.60%questo account1.41%mediana per under 10K

Metrica	Questo account	Mediana per under 10K	Rapporto
Visualizzazioni mediane per post	7 041	4 175	1.69×
Copertura (visualizzazioni ÷ follower)	6.9× audience	2.5× audience	2.77×
Tasso di interazione	0.60%	1.41%	0.43×

Altri account di questa fascia → Confronta con un altro account → Come sono costruiti questi parametri →

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

9.7K4 Sep

66.1K

6.6K

7.5K

5.5K

2.7K

Last 6 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.80%4 Sep

0.68%

0.41%

0.52%

0.48%

1.88%

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

What the audience does

Likes56.3%590 in total

Reposts4.6%48 in total

Replies2.2%23 in total

Quotes1.0%11 in total

Bookmarks35.9%376 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.

Ultimi post

Rispetto ad account della stessa dimensione

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Account simili