LLMS

→ View original post on X — @poe_platform

OpenAI Extends Lead with GPT-5 While Google Gains

By

@poe_platform

–

25 September 2025 19h08

OpenAI extends its lead in text with GPT-5, but Google is gaining ground. OpenAI models now command over 50% of text usage. At the same time, Google’s Gemini-2.5 series has steadily grown to ~10% share, reversing its earlier decline and showing sustained momentum. (4/6)

→ View original post on X — @openai

Claude Opus 4.1 Achieves Expert-Level Performance on GDPval

By

@openai

–

25 September 2025 18h24

On GDPval, expert graders compared outputs from leading models to human expert work. Claude Opus 4.1 delivered the strongest results, with just under half of its outputs rated as good as or better than expert work. Just as striking is the pace of progress: OpenAI’s frontier

LLMs Shaping Politics: Comparing Models with Real-World Polls

By

–

25 September 2025 17h51

The team envisions comparing model outputs w/real-world polls, expanding studies to other democracies, examining newer reasoning or "thinking" models, & exploring downstream effects on users. For now, the dataset helps us understand how LLMs are shaping modern political

LLM Election Data 2024 Dataset Released by Researchers

By

–

25 September 2025 17h51

Authors: @cen_sarah
, @andrew_ilyas
, Hedi Driss, Charlotte Park, @aspenkhopkins
, @charapod
, & @aleks_madry Data: https://
huggingface.co/datasets/sarah
cen/llm-election-data-2024
…

LLMs Political Candidate Trait Association Patterns

By

–

25 September 2025 17h51

Other fascinating findings: – LLMs exhibit strong associations between candidates & traits, like "competent" & "temperamental." When Biden dropped out, words like "charismatic" & "compassionate" moved to Harris, while "competent" & "trustworthy" moved to Trump. – GPT-4 &

MIT team releases 16M LLM responses dataset from 2024

By

–

25 September 2025 17h51

The MIT CSAIL, Sloan, & LIDS team collected each LLM’s responses to their carefully crafted prompts from July to November 2024. They generated a dataset of over 16M time-stamped responses from LLMs, answering questions about topics such as candidate traits, election issues, &

LLM Behavior During 2024 US Election: Stability and Contradictions

By

–

25 September 2025 17h51

The dataset presents a multidimensional view of LLM behavior that captures both stability & change during the first major US election since LLMs became mainstream. It essentially lets you watch these models think, stumble, & sometimes contradict themselves during the 2024

MIT Study: How LLMs Bias Political Election Narratives

By

–

25 September 2025 17h51

How do LLMs respond to questions about the election? MIT researchers ran a 4-month study, querying 12 state-of-the-art models on over 12K prompts daily. They found that the models can reinforce certain political narratives, mislead, or even exhibit manipulative tendencies: