Makes a lot of sense! I should probably stop wondering about this and just put out some interesting academic content then!
13.02.2025 11:59 β π 0 π 0 π¬ 1 π 0@clarkai.bsky.social
PhD student studying AI for tactical decision making in complex ocean environments @uniofbath.bsky.social Distinctly amateur cyclist, triathlete, sailor, climber https://scholar.google.com/citations?user=WE-zZMoAAAAJ&hl=en
Makes a lot of sense! I should probably stop wondering about this and just put out some interesting academic content then!
13.02.2025 11:59 β π 0 π 0 π¬ 1 π 0Does anybody know the level of interaction a general account on BlueSky has? Despite over 1000 followers I feel it's a lot lower than many other social platforms. Do you think this is a result of a large influx of non 'organic' growth?
#meta
#bluesky
#askbluesky
Is there a way to shift my doomscrolling from Instagram to Bluesky? I know I'd end up more informed and positive from 10 minutes on here than on insta. Maybe some good suggestions of random fun accounts to follow?
12.02.2025 21:58 β π 0 π 0 π¬ 0 π 0Wallace and Grommit serving up some great PSA's on AI and automation. Doesn't just seem to play out the AI bad trope and shows to a general audience how it can go wrong quickly.
25.12.2024 22:27 β π 0 π 0 π¬ 0 π 0Starting to debate getting dual providers but can't quite justify the cost. Starlink as back up seems a little steep for redundancy too! So I'm at the whim of virgin media for now....
04.12.2024 21:44 β π 0 π 0 π¬ 1 π 0Oooh I might go for a feeder on the window!
04.12.2024 18:37 β π 0 π 0 π¬ 1 π 0Thank you! Still need to put up my whiteboard! Currently in storage
04.12.2024 18:37 β π 1 π 0 π¬ 0 π 0Yep was very conscious about lighting so put a Philips hue ceiling light in for adjustability and a SAD lamp to fend off the feels
04.12.2024 18:36 β π 2 π 0 π¬ 0 π 0Adorable! But sadly not with my horrific allergies... However a little lamp for the tortoise to come bask under could work
04.12.2024 12:23 β π 2 π 0 π¬ 1 π 0I've built my own desk at a nice sitting height, but was thinking of adding an optional standing riser!
And a full teapot and biscuit tin is non optional
Just starting to build out my home office to have a nice space to push through the end of my PhD. What's your number one item to have in it? Can be anything from plants to monitors!
ππ§ͺ
#phdlife
#academia
All the different kits and bikes before the official change over in the new year! Love seeing the mixes of the new signings integration.
04.12.2024 10:01 β π 2 π 0 π¬ 0 π 0Unfortunately I have broken the illusion of childhood that one can have snacks with no prior work... It has ruined me
01.12.2024 11:40 β π 1 π 0 π¬ 0 π 0Routine is currently, sit and look at my outline in bullet points. Hate most of it, procrastination for the rest of the day π
01.12.2024 11:14 β π 0 π 0 π¬ 1 π 0Great idea giving a talk about it! Might help me figure out the figures I actually want.
01.12.2024 11:11 β π 1 π 0 π¬ 0 π 0Sits down to write paper...
6 months later. This paper hasn't written itself yet no matter how many times I ask nicely...
What are your tips for breaking through writing block and pushing out your paper? Could really do with some advice from the academic community!
π§ͺπ
@carlosduartephd.bsky.social (I hope this is the right @ for you!)
26.11.2024 18:01 β π 1 π 0 π¬ 1 π 0Our oceans aren't quiet and they're getting louder. A great paper on how important sound is in the ocean and how humans are impacting it. This is not only important for the biosphere but louder oceans make bad actors harder to detect.
π§ͺ π
www.science.org/doi/10.1126/...
e introduce the effective horizon, a property of MDPs that controls how difficult RL is. Our analysis is mo- tivated by Greedy Over Random Policy (GORP), a simple Monte Carlo planning algorithm (left) that exhaustively ex- plores action sequences of length k and then uses m random rollouts to evaluate each leaf node. The effective horizon combines both k and m into a single measure. We prove sample complexity bounds based on the effective horizon that correlate closely with the real performance of PPO, a deep RL algorithm, on our BRIDGE dataset of 155 deterministic MDPs (right).
Kind of a broken record here but proceedings.neurips.cc/paper_files/...
is totally fascinating in that it postulates two underlying, measurable structures that you can use to assess if RL will be easy or hard in an environment
For those doing lit reviews, be aware that Scopus can now search for pre prints as well! Great for getting visibility of what's on the horizon in your field.
π§ͺ
Brilliant thank you I'll check it out!
22.11.2024 20:39 β π 1 π 0 π¬ 0 π 0Do you have a link to the whale use index paper? I'm interested at how they account for detection bias.
22.11.2024 19:59 β π 0 π 0 π¬ 1 π 0I've considered doing it but I use an acoustic propagation model in my environment which I think I'd also have to re-write. Wonder if an llm would have a go at refactoring into being jax compatible
22.11.2024 19:49 β π 0 π 0 π¬ 1 π 0Thank you for this starter pack! Would you be able to add me to it? I look at how we deploy and manage ocean sensing platforms for passive acoustic monitoring
22.11.2024 16:32 β π 0 π 0 π¬ 1 π 0These are amazing! Do you have an online store?
22.11.2024 15:26 β π 1 π 0 π¬ 1 π 0The beginnings of a starter pack for UK-based PhD students. Please let me know if you'd like to be added! go.bsky.app/8Uj4vn7
20.11.2024 15:48 β π 1 π 1 π¬ 2 π 0Would love to be added! Great to see so many UK researchers on here
22.11.2024 14:48 β π 1 π 0 π¬ 0 π 0The best intro to RL book there is:
mitpress.mit.edu/978026203924...
What's the performance gains like between gym and gymnax?
22.11.2024 13:38 β π 0 π 0 π¬ 1 π 0What's your reinforcement learning stack?
I use:
- Custom gymnasium env
- Stable baselines 3 - RL library
- PPO - Algorithm
- NatureCNN - Model network
- wandb - run tracking
- wandb sweeps - hyper-parameter optimization
- tensorboard - offline tracking
#AI
#DRL
#RL
#Reinforcementlearning