LLM output can take over your computer
Sounds dramatic, it's pretty fun. And no, this isn't an x-risk thing
inter human agreement
about ai, natural language processing, machine learning, and how these connect with the real human world
Subscribe:.rss.atom.json.md.m3u.pls
Dormant Last read · last published · next check
Read 10 days ago and current, but nothing has been published for 21 months.
Latest episodes
FakeToxicityPrompts: Automatic Red Teaming
Modern LLMs are disappointingly easy to trick into toxicity
Existing threats for an OWASP LLM vulnerability top-n
Shaping LLM security - what vulnerabilities are out there already today, and how can we mitigate them?
The Language Model Vocabulary Gap
How demons creep in, and why magic maybe isn't a terrible metaphor
Kappa scores considered harmful
While popular, this data quality measure has some critical flaws, that have been known for a long time, but are often ignored.
Words for differences in meaning, pronunciation, and spelling
We can describe words that have similar spellings or meanings using special terms. These terms are things like heteronym, homograph…
Four Ways To Spot Fake News Accounts.
Catching lies can be easy, but how can you spot who is paid to tell them?
Pathology of a Fake News Story
How a neutral piece of coverage on the Syrian war is reframed and distributed to misinform the public— a step by step explanation.
Google Scholar Metrics revisited: normalising for publication count
Recently, Google Scholar updated their metrics. There was a bit of surprise in the computational linguistics community, as there is every…