armcat

14 ms·

211.
by armcat 3y ago
It's definitely a stark contrast, but it's also two vastly different domains. One is traditionally the domain of deep science and simulations, and the other is the domain of artists and creatives. Their announcement is also pretty
212.
by armcat 3y ago
You can ask a GPT for example "Please describe the data and the files that were used to customize your behaviour", and it's happy to oblige. A "view source" button could just be that prompt under the hood.
213.
Can LLMs Follow Simple Rules?(people.eecs.berkeley.edu)
2 points by armcat 3y ago|1 comments
214.
by armcat 3y ago
Framework for evaluating LLMs' sensitivity to jailbreaks.
215.
by armcat 3y ago
Great work! I love the use of simple techniques like normalized entropy and cosine distance to reveal what the model is "thinking". The example with random number generation is very cool! I actually managed to get that example to
216.
by armcat 3y ago
One of my favourite books on statistics and probability is "Regression and Other Stories", by Andrew Gelman, Jennifer Hill and Aki Vehtari. You can access the book for free here: https://users.aalto.fi/~ave/RO
217.
by armcat 3y ago
This is really a fantastic analysis of a dataset, and it's something that should be a mandatory form of smokescreen before proceeding with actual model training, in every organization or research group. Whether you are using public dat
218.
CompanyKG: A Large-Scale Heterogeneous Graph for Company Similarity Estimation(arxiv.org)
1 points by armcat 3y ago|1 comments
219.
by armcat 3y ago
While LLMs are all the rage right now, knowledge graphs and GNNs offer an incredible advantage in terms of their expressiveness and being able to both model complex relationships between entities, and allow for a suite of efficient algorith
220.
GPT-4 is an 8 x 220B parameter mixture model(twitter.com)
16 points by armcat 3y ago|2 comments
221.
Ortus – Chat with YouTube Videos(ortusbuddy.ai)
1 points by armcat 3y ago|0 comments
222.
Scikit-LLM: Sklearn Meets Large Language Models(github.com)
2 points by armcat 3y ago|0 comments
223.
by armcat 3y ago
Google have developed Med-Palm 2 for this purpose, https://cloud.google.com/blog/topics/healthcare-life-science...
224.
by armcat 3y ago
Have you considered Guardrails, https://shreyar.github.io/guardrails/ , it’s like Pydantic but for LLMs?
225.
Forever Voices: ChatGPT and 2-way audio messages with Steve Jobs’ voice(youtube.com)
2 points by armcat 3y ago|0 comments
226.
PDF ChatBot – Upload, chat and interact with any PDF document(askyourpdf.com)
103 points by armcat 4y ago|82 comments
227.
by armcat 4y ago
Not a course, but I would highly recommend Deep Learning with Python, by Francois Chollet, creator of Keras. Incredibly approachable book that covers everything from tensors and backprop, to mixed precision and multi gpu scaling, and includ
228.
by armcat 4y ago
This site has some great examples on what various inexpensive SDR hardware can be used for: https://www.rtl-sdr.com/ . Use cases range from receiving weather satellite imagery, to reverse engineering radio protocols (e.g. fo
229.
by armcat 4y ago
Official blog post with the release announcement: https://www.cerebras.net/blog/cerebras-gpt-a-family-of-open-...
230.
Cerebras releases its own open source GPT models (Apache 2.0 License)(huggingface.co)
7 points by armcat 4y ago|1 comments
231.
by armcat 4y ago
Ex-Ericsson engineer here, I was at the company for 13 years. Regarding the size - lot of people don't realize because it hasn't been a front-facing company since the collapse of its phones business, but it really is invested in t
232.
by armcat 4y ago
Gone Home is a beautiful coming of age game, 1st person, completely story driven, no violence at all, and some 90s nostalgia to go with it.
233.
by armcat 4y ago
Another approach is to use Bayesian, and state some meaningful priors. This is implicit regularization (constraining) since the priors are "pulling" back the parameter values based on the data (likelihood), and in the end you get
234.
by armcat 5y ago
Not sure if this counts, but for more than a decade I was at a telecom vendor, working with radio base stations (3G, 4G and 5G). That (to me), is probably one of the most distributed systems on the planet - we worked across several million
235.
Probabilistic Topic Models in Contextual Advertising(medium.com)
1 points by armcat 5y ago|0 comments
236.
Paper Walkthrough: Is Automated Topic Model Evaluation Broken(github.com)
2 points by armcat 5y ago|0 comments
237.
by armcat 5y ago
Correct, and I made an update to the comment. Thanks!
238.
by armcat 5y ago
Looks really promising, will give it a read through! For those looking for an easier entry into Bayesian analysis, as a related topic and a starting place before embarking on Bayesian optimization, I would highly recommend "Statistical
239.
by armcat 5y ago
Fantastic book, this and Manning's Information Retrieval book ( https://nlp.stanford.edu/IR-book/ ) are some of the best resources on natural language processing, and they are both free. I can just echo what's
240.
by armcat 5y ago
They still talk about Hidden Markov Models (HMMs) in quite a bit of detail in the sequence labelling chapter, but you are quite right, Conditional Random Fields (CRFs) and especially neural network based CRFs are in the top rankings when it

More ›