killcoder

6 ms·

1.
by killcoder 4mo ago
I think 'actual parallelism' is a vastly easier and more fruitful way to get better performance out of these kinds of systems, compared to pushing against single-threaded faster generation. Tool calling and responses are often emb
2.
by killcoder 8mo ago
I don't buy the "any constraints cause lower performance via being out of distribution" idea. Sure if you ask the model to output 'reasoning' in JSON steps, that is a completely different 'channel' to its
3.
by killcoder 8mo ago
I was working on a speculative decoding optimisation and its accompanying blog post. Explaining the more basic concepts filled so much of the post I decided to pull them out, forming this article. I had a bit too much fun with the tokenisat
4.
The missed opportunity of constrained decoding(michaelorenstein.com)
2 points by killcoder 8mo ago|2 comments
5.
by killcoder 10mo ago
In South Australia an algal bloom started in ~mid-March of this year, it's a pretty big ecological disaster, probably the worst non-bushfire disaster in living memory. Probably 30% of SA's coastline is affected. It's a pretty
6.
by killcoder 10mo ago
> 30% more token-efficient at the same reasoning level across many tasks But they're claiming it's more token efficient, so me switching my usage to the new model should _free up_ capacity.
7.
by killcoder 10mo ago
It would be nice if users of the codex-cli that are just using API keys as a way to handle rate limits and billing could receive these new models at the same time. I appreciate the reasoning behind delayed 'actual API' release, bu
8.
by killcoder 11mo ago
Apart from the actual model release, this is the second set of models from OpenAI that uses the Harmony response format. I don't suppose anyone knows if OpenAI uses the Harmony format internally for GPT-5 as well? https://co
9.
GPT-OSS-Safeguard(openai.com)
4 points by killcoder 11mo ago|1 comments
10.
by killcoder 2y ago
Not at all, in a renderer the Node and Chromium event loops are bound together, they’re part of the same v8 isolate, no IPC shenanigans. The main process really shouldn’t be used for anything except setup. Since it controls gpu paints among
11.
by killcoder 2y ago
Renderers can access Node APIs via the ‘node integration’ setting or via a preload script.
12.
by killcoder 2y ago
You were correct. Electron lets you expose specific NodeJS APIs via the preload script or everything via the ‘nodeIntegration’ setting: https://www.electronjs.org/docs/latest/api/structures/web-pr... Sep
13.
by killcoder 2y ago
You don’t need IPC, you can either use a preload script to expose particular Node APIs in a secure manner or set ‘nodeIntegration‘ to ‘true’ to expose everything. Source: https://www.electronjs.org/docs/latest/api&
14.
by killcoder 2y ago
Within a renderer you can access NodeJS APIs directly. The main process shouldn’t be used for any significant computation, as it will block GPU paints and cross-process synchronisation. The other main difference is Electron bundles a known
15.
by killcoder 2y ago
Conversely, the last blog post we wrote was 8,000+ words and took months of testing, yet the average 'read' time is under 2 minutes. I'm convinced there's a correlation between interested technical users and the blocking
16.
by killcoder 2y ago
Factorio is a good example of a modern game that works this way.
17.
by killcoder 2y ago
Does DuckDB supports partial reading of .duckdb files hosted externally?
18.
by killcoder 2y ago
I wish this recent backlash was also spent on removing this rootkit. It's a PvE game, I hardly see why a rootkit is warranted.
19.
Driftsort: An efficient, generic and robust stable sort implementation(github.com)
3 points by killcoder 2y ago|0 comments
20.
Overthinking Leetcode's Two Sum with SIMD(darkcoding.net)
2 points by killcoder 3y ago|0 comments
21.
by killcoder 3y ago
We nerd sniped ourselves into testing the latencies of a whole bunch of wireless communication links and protocols for microcontrollers. We’ll probably do a series of power consumption / range tests later on, let me know if there are a
22.
Benchmarking latency across common wireless links for microcontrollers(electricui.com)
135 points by killcoder 3y ago|33 comments
23.
RFC: Rust Has Provenance(github.com)
55 points by killcoder 3y ago|18 comments
24.
by killcoder 3y ago
I work on a product for building user interfaces for hardware devices. All the state management is done via incrementally updated, differential DataFlow systems. The interface is defined in code instead of graphically, but I think that&#x27
25.
by killcoder 3y ago
We've been using (the old) Shiki Twoslash for a few years now for our docs pages: https://electricui.com/docs/components/LineChart https://electricui.com/docs/operators/aggregations
26.
Extracting and Depositing Bits(orlp.net)
3 points by killcoder 3y ago|0 comments
27.
by killcoder 3y ago
WebAssembly in something like wasmtime is probably the closest we've got to this. Though I wish WebAssembly supported more (wider) SIMD instructions.
28.
by killcoder 3y ago
There was a recent post by Voultapher from the sort-research-rs project on Branchless Lomuto Partitioning https://github.com/Voultapher/sort-research-rs/blob/main/wri... Discussion here: https:/&#x
29.
Branchless Lomuto Partitioning(orlp.net)
38 points by killcoder 3y ago|21 comments
30.
A novel branchless partition implementation(github.com)
2 points by killcoder 3y ago|0 comments

More ›