cpldcpu

7 ms·

61.
by cpldcpu 1y ago
Here you can see how i prompted it. I provided a similar paper (also generated with C. opus) as an example, but Opus took it from there: https://claude.ai/share/963b66a7-930c-47a6-a4ea-d7e6993347fa You can find the ref
62.
by cpldcpu 1y ago
Sorry, this is just getting old... Its a trite talking point and not the reason why there are so few consumer-AI companies in Europe.
63.
by cpldcpu 1y ago
But this is just the SFT - "distilled" model, not the one optimized with RL, right?
64.
MiniCPM4 – a series of open multimodal models for edge inference(huggingface.co)
2 points by cpldcpu 1y ago|0 comments
65.
Updates to Apple's On-Device and Server Foundation Language Models(machinelearning.apple.com)
2 points by cpldcpu 1y ago|0 comments
66.
by cpldcpu 1y ago
There is almost no information under the link
67.
by cpldcpu 1y ago
I think that's just a simplifying example. They would most likely not use an image sensor, but a photodetector with a broadband amplifier.
68.
Meta delays release of its 'Behemoth' AI model(reuters.com)
5 points by cpldcpu 1y ago|0 comments
69.
Meta Is Delaying the Rollout of Its Flagship AI Model(wsj.com)
1 points by cpldcpu 1y ago|0 comments
70.
by cpldcpu 1y ago
Is it ARM or x86 based?
71.
by cpldcpu 1y ago
I also strongly suspect that there are earlier sources. However, IRAM looks like compute near memory where they will add an ALU to the memory chip. compute in memory is about using the memory array itself. To be fair, CIM looked much le
72.
by cpldcpu 1y ago
Some more background information: One of the original proposals for in-DRAM compute: https://users.ece.cmu.edu/~omutlu/pub/in-DRAM-bulk-AND-OR-ie... First demonstration with off-the-shelf parts: https:/&#x2F
73.
Matrix-vector multiplication implemented in off-the-shelf DRAM for Low-Bit LLMs(arxiv.org)
230 points by cpldcpu 1y ago|53 comments
74.
by cpldcpu 1y ago
Berlin is not Germany.
75.
by cpldcpu 2y ago
Their approach seems very compelling, but I don't understand if/how they are building a differentiated product? The space of code agents is already pretty crowded.
76.
by cpldcpu 2y ago
That guy leads in with stating that he is missing "autocomplete" in claude code. Cleary a misunderstanding of the scope.
77.
by cpldcpu 2y ago
I thought now everything is about meritocracy? Have we been duped?
78.
by cpldcpu 2y ago
a goof part of team is actually located in europe
79.
by cpldcpu 2y ago
good point, it's certainly more of a trend with the younger generation and in the west. But even living in a country that is perceived as "beer centric", i noticed that people are starting to be much more conscious of their a
80.
by cpldcpu 2y ago
Coming up next and already happening: Alcohol is phased out.
81.
by cpldcpu 2y ago
Thats not really correct. It's actually the beginning of test time scaling. R1 has shown that a very simple reinforcement learning scheme can be used to teach the model how to think in a chain-of-though as an emergent property. No addi
82.
by cpldcpu 2y ago
All this media frenzy around DS V1 makes me feel sick to my stomach. It increased the noise in the AI space by orders of magnitude. Every media outlet is bombarding you with a relentless torrent of half-true information, exaggered interpret
83.
by cpldcpu 2y ago
The $6M that is thrown around is from the DS V3 paper and is for the cost of a single training run for DeepSeek V3 - the base model that R1 is built on. The number does not include cost for personell, experiments, data preparation, chasing
84.
Anthropic CEO Says AI Could Surpass Human Intelligence by 2027(wsj.com)
3 points by cpldcpu 2y ago|3 comments
85.
by cpldcpu 2y ago
These benchmarks are mostly focused on math, which benefits a lot from an improved CoT and is also less sensitive to having "reduced knowledge" in smaller model. Vibes are important in this case...
86.
by cpldcpu 2y ago
I think the main issue is that the average consumer does not know what to do with a raw transformer model. While the base technology is now there and is rapidly improving, a lot of the "glue" and "plumbing" is still miss
87.
by cpldcpu 2y ago
>n the first scenario, an investment in any AI model company that does not own its own compute is like buying a tar pit instead of an oil well. In other words, this future has AI like a commodity and the entity who wins is the one who ca
88.
by cpldcpu 2y ago
>LLMs are awesome but I haven't felt significant improvement since the original GP4 (only in speed). Absolutely disagree. Are you using LLMs for coding? There has been a 10x (or whatever) improvement since GPT4. I causally tracked t
89.
by cpldcpu 2y ago
Voxel rendering is basically raymarching. Current GPUs can implement this easily as pixelshaders. Plenty of examples on https://www.shadertoy.com/
90.
by cpldcpu 2y ago
yeah, the title is technically correct. but...

More ›