🚀 By Science and FOR SCIENCE 🇫🇷

PhysiQuanty PRO

PhysiQuanty

·

PhysiQuanty

AI & ML interests

Theoretical Physics 🇫🇷 The perceptron is not a neuron ; ElectroChemical Calculations (NMDA & AMPA receptor, apical and basal dendrite, axon initial segment, metaplasticity, ... )

Recent Activity

liked a dataset 3 days ago

WhirlwindAI/Binance-PERP-Data-Multi-TF

liked a dataset 16 days ago

Bayway/italian-open-timeseries

reacted to Bayway's post with 🧠 16 days ago

Italian Open Time Series: 46 public Italian series for zero-shot forecasting, and a first live test of Timestral Italian data is almost absent from the corpora that time-series foundation models are trained and benchmarked on. So I collected 46 open series (CC0/CC BY): smog in Milan, bikes in Bologna, tourism, energy, earthquakes, deaths, Italian Wikipedia. Gaps stay NaN, and a tasks.yaml makes every window reproducible with fev. Seven models, zero-shot, 19 tasks (Timestral 5M to 209M, Chronos-2, TimesFM 3.0, Toto 2.0). Skill vs seasonal naive, fev-bench style: • Backtest, windows ending 14 Oct, Timestral's release day: Toto 2.0 28.4, Timestral-209M 28.2 • Live, data published after that day (6 tasks, ~1 week): Timestral-209M 32.2, Timestral-123M 30.9, TimesFM 3.0 30.3, Toto 2.0 27.7 • Monthly consumer prices of electricity and gas: plain AutoETS beats every foundation model The live round is small; I'll re-run it every week. Two surprises. Weather beats the calendar. On a Bologna bike counter, giving Timestral the observed weather of the forecast days (a perfect forecast) cut the error by 17%. Holidays and bridge days alone: 4%. Daylight saving time. The model works in UTC; cyclists don't. Across the 25 Oct clock change my first version put Monday's rush hour at 5 am, until I moved the series to Italian wall-clock time. For Bologna there's a small page: pick a street and a day in the next week for an hourly bike forecast, computed live from the city's counts and Open-Meteo, plus a map of counters and air-quality stations. Dataset: https://huggingface.co/datasets/Bayway/italian-open-timeseries Bologna: https://huggingface.co/spaces/Bayway/due-torri-due-ruote Not included: GME wholesale power prices and Terna load data. Their terms forbid redistribution, and GME bans automated extraction for any purpose, AI training included.

View all activity

Organizations

PhysiQuanty 's models 13

PhysiQuanty/Self-Predicting-Gradient-Descent

PhysiQuanty/Binary-LLM-POC

Text Generation • 10.7M • Updated May 2 • 255 • 12

PhysiQuanty/Wiki-Test2

79.7M • Updated Mar 4 • 10

PhysiQuanty/Wiki-Test

79.7M • Updated Mar 3 • 14 • 1

PhysiQuanty/Patenty-Test2-Radix-65536

79.7M • Updated Mar 2 • 14

PhysiQuanty/VOCAB-4294967296-FOUR-LOGITS-256

13.7M • Updated Feb 25 • 17 • 2

PhysiQuanty/Patent-Dual-Cross-Entropie

46.5M • Updated Feb 24 • 5

PhysiQuanty/Patent-Test-Radix-65536-AutoTokenizer_FineTune

79.7M • Updated Feb 23 • 22

PhysiQuanty/Patenty-0.1

79.7M • Updated Feb 22 • 6

PhysiQuanty/Patenty1-0.1B

79.7M • Updated Feb 19 • 4

PhysiQuanty/Patent-Test-Radix-65536-AutoTokenizer

79.7M • Updated Feb 13 • 9

PhysiQuanty/Patent-Test-Radix-65536

79.7M • Updated Feb 12 • 7

PhysiQuanty/Binary-Addition-LLM-POC

Text Generation • 10.7M • Updated Feb 5 • 15 • 4