BTC $83,315.99 -1.68%
ETH $2,672.29 -0.67%
BNB $764.43 -1.76%
XRP $1.48 -3.39%
SOL $118.15 -4.04%
TRX $0.3361 +0.71%
DOGE $0.0932 -4.15%
ADA $0.2442 -4.51%
BCH $307.07 -8.28%
LINK $14.94 +5.99%
HYPE $87.71 -4.49%
AAVE $146.57 -5.53%
SUI $1.14 -9.13%
XLM $0.2277 +5.04%
ZEC $1,456.93 -9.51%
AAPL $338.31 -0.61%
AMZN $246.23 -1.60%
GOOGL $342.19 -0.62%
MSFT $509.48 -1.55%
META $717.13 -4.26%
NVDA $228.61 +1.40%
TSLA $358.45 -4.01%
SNDK $1,711.08 -4.09%
INTC $115.89 -7.85%
SPCX $146.52 -1.59%
MU $1,054.24 -3.93%
AMD $606.77 -4.37%
BTC $83,315.99 -1.68%
ETH $2,672.29 -0.67%
BNB $764.43 -1.76%
XRP $1.48 -3.39%
SOL $118.15 -4.04%
TRX $0.3361 +0.71%
DOGE $0.0932 -4.15%
ADA $0.2442 -4.51%
BCH $307.07 -8.28%
LINK $14.94 +5.99%
HYPE $87.71 -4.49%
AAVE $146.57 -5.53%
SUI $1.14 -9.13%
XLM $0.2277 +5.04%
ZEC $1,456.93 -9.51%
AAPL $338.31 -0.61%
AMZN $246.23 -1.60%
GOOGL $342.19 -0.62%
MSFT $509.48 -1.55%
META $717.13 -4.26%
NVDA $228.61 +1.40%
TSLA $358.45 -4.01%
SNDK $1,711.08 -4.09%
INTC $115.89 -7.85%
SPCX $146.52 -1.59%
MU $1,054.24 -3.93%
AMD $606.77 -4.37%

xpla

All
Article
Flash

Tether released a dataset of 191.4 billion Token STEM, betting on the "explanatory ability" of local small models

The stablecoin giant Tether's Tether AI Research today released QVAC Genesis III, a synthetic training dataset aimed at the fields of science, technology, engineering, and mathematics, with a scale of 19.143 billion tokens, covering 159.6 million documents.It is reported that Tether hopes to enhance the reasoning and teaching capabilities of small AI models, allowing more AI assistants to run directly on laptops, mobile phones, and local servers, reducing the ongoing reliance on large cloud models.Genesis III continues the data lineage of Genesis I and Genesis II, covering 19 curriculum-related fields including biology, chemistry, physics, mathematics, computer science, medicine, electrical engineering, and machine learning, aimed at high school, university, and professional applications. Tether states that the dataset will focus on training models to explain problem-solving paths, identify erroneous reasoning, and provide corrections, rather than just outputting final answers.According to Tether's disclosed test results, a 1.7 billion parameter model trained with Genesis III option-level data achieved an effective answer rate of 99.45% in relevant benchmark tests. Compared to the Cosmopedia-v2 training model of similar token scale, Genesis III improved scores on the ARC-Easy, ARC-Challenge, and MMLU STEM benchmarks by 28.57, 21.35, and 15.03 percentage points, respectively.

"The Fed's Mouthpiece": How Waller Explains Inflation Determines the Path He Will Take

According to Nick Timiraos, a reporter for the Wall Street Journal, known as the "Fed's megaphone," in his latest article, Federal Reserve Chairman Kevin Walsh will face a core issue during his first major speech this week in Jackson Hole: whether the persistently high inflation in the United States is caused by one-time shocks such as tariffs and wars, or if the economy itself is still overheating.This judgment will directly determine the direction of interest rates and is currently the biggest divergence within the Federal Reserve. In the July meeting, three officials supported a rate hike, while other officials indicated the possibility of further tightening policies, but Walsh has not made a clear statement. Since taking office, he has deliberately reduced policy guidance, and now both the market and his Federal Reserve colleagues are waiting for him to systematically explain his judgment for the first time.The key to Walsh's term ultimately depends on how he explains why the previous policies failed to bring inflation back to 2%. If the policies of lowering interest rates and supporting employment over the past two years were fundamentally wrong because the labor market is actually stronger than the Federal Reserve judged, then Walsh needs to push for a reversal of interest rate cuts. However, this would conflict with the previous demands from Trump and Bessenet for further rate cuts.
app_icon
ChainCatcher Building the Web3 world with innovations.