Ohio State University

Headlines · 4

1. Oct 23
   AI megafunds reshape venture capital ahead of major IPOs

   Newcomer reports that Anthropic’s pre-IPO fundraising prompted Menlo Ventures, Lightspeed and ICONIQ to raise larger funds, seek co-investors and market special-purpose vehicles to write bigger checks. OpenAI is reportedly discussing a $30 billion fundraise at a $1.25 trillion valuation, which Newcomer attributes to CNBC. The analysis also says a Stanford and Ohio State ranking of 50 AI-focused venture firms placed a16z first, ahead of Lightspeed, Menlo Ventures, Sequoia and ICONIQ Growth.

   Newcomer · 🔥 0
2. Oct 23
   Scenttronic prototype transmits five odor classes over a network

   According to the supplied report, Ohio State University researchers published a laboratory prototype called Scenttronic that digitizes odors with an AI electronic nose and transmits them through a cloud module to wearable receivers. In a blind test involving 16 volunteers, participants distinguished five odor classes: cake, coffee, orange, red wine and beer. The wearable receiver uses a five-channel piezoelectric atomizer to release preconfigured scent sources and must be held close to the wearer’s face. The report says the system remains a laboratory prototype, supports few odors, and has limited accuracy for complex mixtures. It does not provide a numerical accuracy rate for the blind test, and the prototype replays configured scents rather than generating arbitrary odors.

   cnBeta · 🔥 0
3. Oct 21
   Ohio State's AI E-Nose Sends Smells to a Wearable Scentronic Receiver

   Researchers at Ohio State University have developed a system that detects an odor with an AI-powered electronic nose and transmits the scent data to a receiver called Scentronic, which recreates the smell, according to a study released Wednesday and reported by CNET. The team built wristband, handheld and necklace versions of Scentronic, each with a five-channel scent distributor using a piezoelectric atomizer and powered by Wi-Fi and a cloud server that stores and transfers the scent data. In a test with 16 participants, the system transmitted five scents — cake, beer, red wine, coffee and oranges — and the testers accurately identified all of them; lead author Chunyu Yang said the gas-sensor array combined with machine learning allows a compact sensing system that can be integrated with wearables.

   MIT科技评论中文 · 🔥 0
4. Oct 5
   TRACE Aligns FP4 Training and Rollout for MoE Reinforcement Learning

   Alibaba Qwen and Ohio State University researchers proposed TRACE, a framework that uses rollout-side FP4 quantization outcomes to guide training-side rounding for reinforcement learning of MoE language models. Evaluated on four models across reasoning, coding, and long-horizon RL tasks, the authors report performance comparable to BF16 rollout with joint FP4 weights, activations, and KV cache, plus up to a 5.4× rollout speedup. A compact caching scheme stores selected mantissa and scale information to reduce the overhead of transferring rollout guidance.

   Hugging Face · Papers · 🔥 0

Experience and discussion from the community

Share my Ohio State University experienceAsk about Ohio State University