Trend Engine: AI-Powered News & Trends

Best Aesthetic things dropship supplier in India

not sure what the spikeyness is caused from… was thinking weekends maybe?

FX vol surface for EURUSD above

Hi all

Iv been developing my own stratergy and completed (they are never complete right?) my engine and deployment system.

My strategy shows good promise but is fully technical (loosely based around opening range, RVOL and technical sentiment / daily bias)

I’m looking to throw market sentiment into the mix and see if I can add to my directional bias to sharpen confluence.

I’m potentially looking to gather news scoring on ticker level and looking to create a weighted moving average to sentiment score, short term due to ORB frequency, perhaps 7 days weighted.

Can anyone recommend if this is a good / typical approach?

Can anyone recommend and data sources? I’m looking at market aux at the moment, any good?

Ideally it would be nice to get some free data for a couple of years, a couple of tickers so I can prove concept before paying for data, delay is fine as it’s only for back testing – if anyone has this data to hand for a ticker or 2 I would appreciate a share just for testing (not being tight, I just dont want to pay for a sub for a conceptual idea)

Longer term, my system uses around 15 tickers but I have collected detailed spread and 8 years of 1m data for around 50 tickers so if it shows promise I would like to interfere on all of the tickers for testing.

Thanks.

I’ve posted on this sub before, but context is that me and a small team are working on a [benchmark](https://www.designarena.ai/) to evaluate how good LLMs are at producing UIs and frontends that are engaging and satisfiable for people.

Right now, working on adding more models, and specifically open source models developed by individual developers (or a small group of developers). Above is the current top 10 in the leaderboard. If you’re interested, just send me a DM.

Here are some requirements:

1. Inference needs to be fairly quick (max should take 3 minutes on average). Models are writing html/css/js code on the order of 4K-10K tokens on average.
2. Give us a logo and name for the provider/org you want the model to be associated with
3. An api endpoint that we can call with your desired parameters for the model. It needs to ideally be able to support a few concurrent requests at a time and around \~500 requests a day (though you can rate limit us if you would like to cap it at a smaller number)

Scaling language models unlocks impressive capabilities, but the accompanying computational and memory demands make both training and deployment expensive. Existing efficiency efforts typically target either parameter sharing or adaptive computation, leaving open the question of how to attain both simultaneously. We introduce Mixture-of-Recursions (MoR), a unified framework that combines the two axes of efficiency inside a single Recursive Transformer. MoR reuses a shared stack of layers across recursion steps to achieve parameter efficiency, while lightweight routers enable adaptive token-level thinking by dynamically assigning different recursion depths to individual tokens. This allows MoR to focus quadratic attention computation only among tokens still active at a given recursion depth, further improving memory access efficiency by selectively caching only their key-value pairs. Beyond these core mechanisms, we also propose a KV sharing variant that reuses KV pairs from the first recursion, specifically designed to decrease prefill latency and memory footprint. Across model scales ranging from 135M to 1.7B parameters, MoR forms a new Pareto frontier: at equal training FLOPs and smaller model sizes, it significantly lowers validation perplexity and improves few-shot accuracy, while delivering higher throughput compared with vanilla and existing recursive baselines. These gains demonstrate that MoR is an effective path towards large-model quality without incurring large-model cost.

I keep re-evaluating Web 3.0 every few months to check if things have become easier in this space. Brave was the easiest in this area, but I do remember having installed web extensions in Firefox, which apparently seem to have disappeared? The only extensions they do provide today are for Google Chrome. Am I missing something very obvious, or is this really how it is?

Tight range in daily chart of bitcoin. I am buying