Finally watched Ramayana, and I genuinely feel like giving a reality check to everyone questioning the casting and criticizing the so-called glorification of Ravana.
For the first time, a film has properly explored who Ravana really was, how he became so powerful, and where his extraordinary strength and abilities came from.
Showing Ravana's power doesn't mean glorifying him! We have always seen Lord Rama's greatness and divine strength, but that doesn't mean Ravana wasn't powerful. Perhaps some people are simply unable to accept that a villain can be extraordinarily powerful too.
For me, this is the best portrayal of Ramayana since Ramanand Sagar's iconic Ramayan.
The casting, performances, storytelling, and absolutely stunning VFXโeverything is phenomenal!
A true masterpiece! ๐ฅ๐ I highly recommend everyone watch it before judging. The makers absolutely nailed it! โฅ๏ธ
Are we really doing justice to AI? ๐ค
Are we using AI to improve our skills, creativity, and critical thinking, or are we simply taking shortcuts to achieve success?
AI should empower human intelligence, not replace the effort required to develop it.
Is AI making us smarter, or just more dependent on it?
What do you think? ๐ญ
Weโre noticing an interesting trend where later ICLR submissions tend to receive lower scores. I hypothesize that many reviewers might be relying on ChatGPT to evaluate papers. Since ChatGPT is already aware of ICLRโs historical acceptance rates, it might implicitly bias its evaluations. As more papers are reviewed, this could potentially influence the scoring pattern, resulting in lower scores for later submissions. Of course, this is just a hypothesis, but it could be one possible explanation for the trend weโre observing. ๐ฅฒ #ICLR2027
@mbilgehanertan This trend might be explained by reviewers relying on ChatGPT to evaluate papers. My hypothesis is that ChatGPTโs knowledge of ICLRโs historical acceptance rates could introduce unintended bias into its evaluations, potentially contributing to lower scores for later submissions.
Iโm very happy to share that our paper โEntropy-Gated Latent Recursionโ has been accepted to NeurIPS 2026! ๐
This work was a great collaborative effort with Soham Bhattacharjee, @salemlahl , @TakacMartin , and @NilsLukas7 . Iโm grateful to the entire team for their contributions and support throughout the project.
Looking forward to sharing the work and meeting the community at NeurIPS 2026!
๐ Entropy-Gated Latent Recursion: https://t.co/oWplPsa3YZ
In Advances in Neural Information Processing Systems (NeurIPS), 2026.