Least-Squares Concept Erasure with Oracle Concept Labels
23 May

Least-Squares Concept Erasure with Oracle Concept Labels

stp2y0 CommentsGenAI
This post assumes some familiarity with the idea of concept erasure and our LEACE concept erasure method. We encourage the reader to consult our arXiv paper for background. For a PyTorch implementation of this method, see the OracleFitter class in our GitHub repository. WARNING: Because this erasure transformation depends on the ground truth concept label, it can increase the nonlinearly-extractable information about the target concept inside a representation, even though it eliminates the linearly available information. For this reason, optimizing deep neural networks on top of O-LEACE'd representations is not recommended; for those use cases we recommend vanilla LEACE. In…
Read More
Sigma Secures $200M Round to Advance Its BI and Analytics Solutions
23 May

Sigma Secures $200M Round to Advance Its BI and Analytics Solutions

stp2y0 CommentsViral News
(NicoElNino/Shutterstock) Sigma Computing, a cloud-based analytics solutions provider, has raised $200 million in Series D funding to further advance its efforts in broadening BI use within organizations by enabling users to query and analyze data without writing code.  The latest rounding of funding takes the vendor’s total funding to $581.3 million with a valuation estimated to be around $1.5 billion, a staggering rise of 60% since the last funding round in 2021. The steep rise in valuation is partially a result of rising demand for greater productivity and monetization in the era of cloud data transition.  Spark Capital and Avenir…
Read More
Why ChatGPT feels more “intelligent” than Google Search
23 May

Why ChatGPT feels more “intelligent” than Google Search

stp2y0 CommentsNews
The Google Search bar doesn’t feel like an artificial intelligence. No one speculates that it might soon become an artificial general intelligence (AGI) — an entity that is competitive across many domains to a human being.But do you know many “generally intelligent” humans who can muster a decent translation into and out of 133 languages? Or a co-worker who can perform mathematical operations instantly, know a good route between pretty much any two locations on the planet, and proffer a plausible answer to every question you might ask, in less than a second? Sure, these answers aren’t original, and nor are they…
Read More
LeoLabs zeroes in on anomalies in satellite operations
23 May

LeoLabs zeroes in on anomalies in satellite operations

stp2y0 CommentsChat-GPT
COLORADO SPRINGS – LeoLabs, the Silicon Valley startup mapping activity in low-Earth orbit, is relying on artificial intelligence to spot anomalous satellite operations. A LeoLabs visualization tool shown at the 39th Space Symposium tracks maneuvers performed by satellites that change their orbits frequently. And it highlights maneuvers conducted by satellites that did not typically perform them. Three of the first satellites in a Chinese communications constellation that could include 12,000 satellites, for example, remained in stable orbits for months after they were launched in late 2023. “Then at the same time, all three executed an organized maneuver campaign,” Owen Marshall,…
Read More
AT&T CEO: Focused on Always-On Connectivity
23 May

AT&T CEO: Focused on Always-On Connectivity

stp2y0 CommentsRAG models
AT&T CEO John Stankey joins Caroline Hyde and Ed Ludlow following his comments on the company's multi-year growth strategy at the JPmorgan technology conference. He discusses slowing demand for handsets, always-on connectivity, and why he thinks AST is more "consumer-centric" than Starlink. He speaks on "Bloomberg Technology." (Source: Bloomberg) Source link lol
Read More
Data Machina #251
23 May

Data Machina #251

stp2y0 CommentsViral News
Three New Powerful Open AI Models. I’m told by colleagues at Hugging Face that just a week since LLama-3 was released, more than +10,000 model derivatives have been developed! The pressure on black-box, closed AI models is huge, and achieving GPT-4 performance with open, smallish models is upon us. Which is great. In the last few days, three new, smallish, powerful open AI models were released. Interestingly enough, the power of these 3 models is based on a combination of: 1) Innovative training architectures and optimisation techniques, and 2) Data quality for different types of data (synthetic, public or private).…
Read More
23 May

Consultant behind deepfaked Biden robocall fined $6m for Democratic primary scheme

stp2y0 CommentsChat-GPT
Steve Kramer, a political consultant who admitted that he deepfaked Joe Biden’s voice in a robocall that was sent out to thousands of US voters in January 2024, has been indicted and fined $6m.The robocall, which went out ahead of the first Democratic presidential primary in the US in New Hampshire, used artificial intelligence to fake Biden’s voice telling voters to stay home and “save” their votes for the November general election.Ten criminal charges were filed against Kramer out of Rockingham County on 22 May, including allegations of bribing, intimidation and impersonation of candidates, TV station WMUR in New Hampshire…
Read More
23 May

Thoughts on Academia and Industry in Machine Learning Research • David Stutz

stp2y0 CommentsGenAI
Introduction By construction, a PhD has a clear end. Depending on the program, country and field, a PhD is supposed to be done within 3-6 years when it is usually awarded after an official defense of the research work. This is in contrast to most other careers and jobs, especially in industry but also in the public sector. Even though a PhD is often considered as a qualification for independent research and thereby acts as the entry to an academic career, it is commonly assumed that most PhD graduates do not continue in academia. This also matches my impression and…
Read More
No widgets found. Go to Widget page and add the widget in Offcanvas Sidebar Widget Area.