Loading activity…
@srchvrs
Machine learning scientist and engineer @Oracle speaking πtorch & C++. Past @awscloud, @LTIatCMU. Opinions sampled from MY OWN 100T param LM.
AI
The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.
AI
Futurism, citing Bloomberg, reports most Nova models remain supported with minimal resources as Amazon shifts to a frontier-model effort.
AI
In an earlier post, @lucy3_li described the research as examining unequal treatment of English varieties in high-stakes settings.
AI
A user says the keynote covers choosing important problems, making an impact and how many EMNLP papers to publish.
AI
OpenAI says its claimed refutation of the Exponential Time Hypothesis was formalized in Lean and made public.
AI
A user linked to an arXiv paper they described as fixing the issue; another user praised the work.
XAI
The poster contrasts that with METR's reported three-hour horizon, while noting real-world R&D tasks may be harder.
AI
Bhuwan Dhingra says the tool handles math, but AI-polished writing can also be flagged.
XAI
A user asks how models could learn to do this and points to a thread.
AI
A user pairs the datacenter quip with a photo, without adding context in the text.
AI
The essay’s author shares personal reflections on a growing disconnect between scientific rigor, professional recognition and organizational rewards.
AI
A member of the paper’s team says a little fine-tuning takes the models further.
AI
A commentator disputes AI 2027's human-response forecasts, arguing labs accelerate when perceived risk falls and act when it rises.
AI
One poster calls Anthropic early on reinforcement learning and tool use; a reply points to a LaMDA paper.
AI
A user calls something “the future of coding agents” but does not identify what they mean.
AI
A user frames the choice around how an AI’s intelligence compares with their own.
AI
The author says AutoRubric's 25-recipe cookbook includes runnable Python examples on judge calibration and cost control.
AI
The paper's authors say the Lean-verified formal proof does not correspond to OpenAI's announced natural-language Navier-Stokes argument.
AI
Peter Norvig argues that probabilistic language models can capture usage patterns that simpler grammatical rules miss.
AI
A user says models wrote a couple hundred lines of Blender Python per scene, arranging cameras, lighting and materials.
