Researchers find you don’t need a ton of data to train LLMs for reasoning tasks
With a few hundred well-curated examples, an LLM can be trained for complex reasoning tasks that previously required thousands of instances.
decrypt1d
New Open Source AI Model Rivals DeepSeek's Performance—With Far Less Training Data
OpenThinker-32B achieved benchmark-beating results using just 14% of the data its Chinese competitor needed, marking a win ...
Techopedia4d
Diffbot’s AI Model Suggests “Smaller Is Better” for LLMs
Learn whether a smaller Diffbot’s AI model with an innovative GraphRAG AI training technology can solve AI hallucinations for ...
HackerRank Introduces New Benchmark to Assess Advanced AI Models
Industry Leader Known for Software Development Skills Expertise Introduces Real-World Benchmark of AI Software Development CapabilitiesCUPERTINO, Calif., Feb. 11, 2025 (GLOBE NEWSWIRE) -- HackerRank, ...
Digital information world5h
End of an Era? CPU Performance Drops in 2025—What’s Really Happening?
According to new data from PassMark Software, which has offered PC benchmark testing tools since 1998, average CPU ...
Techopedia1d
Kimi AI 1.5: New Chinese AI Model Beats ChatGPT & DeepSeek
Just days after DeepSeek R1 made headlines, Moonshot AI introduced Kimi AI 1.5, a model already touted superior to OpenAI’s ...
Is Perplexity's Sonar really more 'factual' than its AI rivals? See for yourself
The company claims its newly upgraded model is number one in user satisfaction and speed - but its methodology is unclear.
Hosted on MSN1d
Benchmark Initiates Coverage of Tesla (TSLA) with Buy Recommendation
Fintel reports that on February 12, 2025, Benchmark initiated coverage of Tesla (NasdaqGS:TSLA) with a Buy recommendation.
Hong Kong stocks hit 4-month high as DeepSeek-driven revaluation takes hold
The rally in Chinese stocks spurred by DeepSeek’s AI model still has some way to go, according to UBS and Morgan Stanley.
Which AI agent is the best? This new leaderboard can tell you
On Wednesday, Galileo launched an Agent Leaderboard on Hugging Face, an open-source AI platform where users can build, train, access, and deploy AI models. The leaderboard is meant to help people ...
OpenAI’s DeepResearch can complete 26% of ‘Humanity’s Last Exam’ — a benchmark for the frontier of human knowledge
OpenAI’s o1 and DeepSeek’s R1 models, which previously sat atop the leaderboard, could only get through roughly 9% of the ...
Hosted on MSN1d
Jordan-Qatar Labor Ties Set Benchmark For Arab Cooperation - Minister
Minister of Labor Khaled Bakkar highlighted Qatar-Jordan relations as an exemplary model for Arab economic cooperation, reflecting deep-rooted bilateral ties and the historic relationship between both ...