nmitchko

11 dni temudiscuss

Poster: The cool part is that the model literally talks itself out of giving the cached wrong answer with a latent only thinking budget.

nmitchko

3 miesiące temudiscuss

A fantastically simple solution to improving algorithms, I wish I had this years ago in activation engineering: https://blog.n.ichol.ai/llm-activation-engineering-an-easy-f...

How do I access AlphaEvolve?

nmitchko

4 miesiące temudiscuss

I used those two in combination to fix pain after 3x surgeries to repair a torn pec + infection. They work and helped me heal from being at a 3/10 constant pain down to baseline.

Not something I would do at any point for fun. But anecdotally, it's materially better than other alternatives offered/available.

nmitchko

5 miesięcy temudiscuss

Can someone make a startup that allows me to do this as an individual?

nmitchko

10 miesięcy temudiscuss

In case anyone wants to do this themselves, check out the pipeline here: https://github.com/isc-nmitchko/iris-document-search

Colnomic and nvidia models are great for embedding images and MUVERA can transform those to 1D vectors.

nmitchko

10 miesięcy temudiscuss

Next steps for AI in general:

  - additional modalities
  - Faster FPS (inferences per second)
  - Reaction time tuning (latency vs quality tradeoff) for visual and audio inputs/outputs
  - built-in planning modules in the architecture (think premotor frontal lobe)
  - time awareness during inference (towards an always inferring / always learning architecture)

nmitchko

2 lata temudiscuss

No other models that are public worth comparing to... Hippocratic advertises good benchmarks but that might be marketing fluff.

Have you checked out dataset building with nemotron? The nemotron synthetic data builder is quite powerful.

Moreso, check out model merging. It's possible if you merge some of your model against llama3.1 base it may perform much better.

Check out max labonne's work on hugging face

nmitchko

2 lata temudiscuss

Interesting they don't compare to open-bio. Page 7 charts are quite weak.

https://huggingface.co/aaditya/Llama3-OpenBioLLM-70B

nmitchko

2 lata temudiscuss

We're excited to share pitchpilot with the HN community. Our beta users have found the embedded audio particularly useful for enterprise sharing. We're keen to keep improving, and our mission is to make communication easier.

In the roadmap is adding video export, digital twin presentations, and real-time presentations. We don't wrap a public LLM, so we don't share any data.

nmitchko

2 lata temudiscuss

Given that Generative AI can now read brain scans [1] and this, I wonder how far away we are from "you thought negatively about something, the authorities are on their way".

[1] -- https://www.biorxiv.org/content/10.1101/2022.11.18.517004v3

nmitchko

2 lata temudiscuss

Tin-foil hat time:

1. First, models will predict pollution. The outcomes will help shape urban policy. But these won't solve crime or stop people from driving.

2. Second, models will predict individual behavior and track person level emissions. The outcomes will force behavior changes, mostly freedom limiting.

3. Third, and finally, models will predict thoughts. The the thought of driving instead of walking might trigger a response.

It's a slippery slope and we need to draw a line between prediction and policy.

nmitchko

2 lata temudiscuss

How does this compare to ehealthexchange or other qhins that have many years of experience and charge lower costs?

nmitchko

2 lata temudiscuss

It truly feels like the space race in terms of building LLMs right now. Question is, who lands on the moon first?

nmitchko

2 lata temudiscuss

This reminds me of the matrix movie scene when they look at the encrypted thoughts of the matrix.

https://cdn.swisscows.com/image?url=https%3A%2F%2Fi.pinimg.c...

nmitchko

3 lata temudiscuss

Great work, will try this tonight.

Only question, why do you name variables with the λ symbol?

nmitchko

3 lata temudiscuss

BTC =/= crypto at large.

Despite all the bad press, shady exchanges, bubbles and busts, a large amount of BTC owners don't sell, and that alone will drive it's price up. Other cryptos, I can't speak for.

nmitchko

3 lata temudiscuss

California politicians confused why large companies continue to mass-migrate to Texas

nmitchko

3 lata temudiscuss

Unfortunately, the NSA & NIST most likely is recommending a quantum-proof security that they've developed cryptanalysis against, either through high q-bit proprietary technology or specialized de-latticing algorithms .

The NSA is very good at math, so I'm be thoroughly surprised if this analysis was error by mistake rather than error through intent.

nmitchko

3 lata temudiscuss

With all we see around cancer, it's time to study fembendazole and other anti-tubulins seriously.

nmitchko

3 lata temudiscuss

For each step in the article:

1. You should really download the huggingface hosted model.

2. Why convert the model if meta already hosts it here: https://huggingface.co/meta-llama

3. This step completely glossed over the hardware requirements. Doesn't explain any of the instructions needed in the finetuning process.

4. "Now run the model". What? What about the model precision, hosting, standard open source tools, etc,etc...

Submissions

The Doctor Is Not the Mother: DS4 Latent Reasoning

Show HN: We built pitchpilot, an AI synthesizer to tailor presentations

comments