What is happening in Pakistan? 200% hike on high octane fuel, 20% hike on petrol and diesel happened overnight. Petrol now is sold at 321 PKR a litre. Smart lockdowns announced in Sindh province, so that their fuel can be conserved. Restricting movement, gatherings and public events. Schools are shut for two weeks. Government offices have moved to four day working week. Private offices told to shift 50% staff to work from home. We're not doing any of it. Still some leaders are spreading rumours that there will be lockdown. This rumour mongering should not happen. It's being done to spread fear. Markets and shopping centres are ordered to be closed by 9.30 PM in Pakistan. These are all the media reports. All universities shut down and shifted to online learning in Bangladesh as there's no electricity. Five hour rotational power cuts are happening. Implemented for domestic consumers in the city of Dhaka. Fuel station across Dhaka closed due to shortage of octane and diesel. So, the situation across the world is not good. We're managing in a way that our citizens don't face any difficulty. There's a contrast in price I want to say. In India, nothing has changed. Whereas in Pakistan, plus 20% to plus 200% price depending on petrol, diesel, high octane etc. Bangladesh has rationed stations closed, supply cut 10 to 15%. Excise duty action, we have cut Rs. 10 per litre. Neither Pakistan nor Bangladesh have responded. So, India among these neighbourhood countries is maintaining a level of stability. We're following Hon'ble PM Shri @narendramodi Ji's guidance on the same. - Smt @nsitharaman in Rajya Sabha

The world is moving towards agents Static benchmarks don't measure what agents do best (multi-turn reasoning) Thus, interactive benchmarks: * Terminal Bench (@alexgshaw, @Mike_A_Merrill) * Text Arena (@LeonGuertler) * BALROG (@PaglieriDavide, @_rockt) * ARC-AGI-3 (@arcprize)

We found that agents generate progressively worse code with each iteration. Real developers do not. SlopCodeBench is the only eval that faithfully measures quality degradation on iterative, long-horizon coding tasks. arxiv.org/abs/2603.24755 scbench.ai 🧵c