DeepReinforce (@deep_reinforce) 's Twitter Profile
DeepReinforce

@deep_reinforce

Trialing and reinforcing the path to tomorrow.

ID: 1940143597506142208

linkhttp://github.com/deepreinforce-ai calendar_today01-07-2025 20:21:02

31 Tweet

73 Takipçi

98 Takip Edilen

DeepReinforce (@deep_reinforce) 's Twitter Profile Photo

On Anthropic’s take-home challenge, IterX has achieved 1,133 cycles using its deep thinking model without any human knowledge, outperforming the best Claude 4.5 version (1,363 cycles) , strong enough to be considered hired by Anthropic. Training is still ongoing, stay tuned.

On Anthropic’s take-home challenge, IterX has achieved 1,133 cycles using its deep thinking model without any human knowledge, outperforming the best Claude 4.5 version (1,363 cycles) , strong enough to be considered hired by Anthropic. 
Training is still ongoing, stay tuned.
DeepReinforce (@deep_reinforce) 's Twitter Profile Photo

Finally got a chance to hold the leaderboard of Anthropic's take-home challenge for just a second, achieved with the help of Iterx using RL

Finally got a chance to hold the leaderboard of Anthropic's take-home challenge for just a second, achieved with the help of Iterx using RL
DeepReinforce (@deep_reinforce) 's Twitter Profile Photo

🥳🥳 Demo on using IterX to achieve ~1140 cycles on Anthropic's take-home challenge without trouble shooting. 😀😀Feedback is deeply appreciated !!