not-a-cat

letzten Monatdiscuss

This wouldn't happen if employees rejected cash-based compensation and decided to be founders themselves. Most employees trade risk for higher cash comp, and end up with less upside. This issue is mostly settled by the employment market

not-a-cat

vor 4 Monatendiscuss

Tufalabs | Senior research/engineers | Full Time | Zurich, Onsite | 100-200k CHF | https://tufalabs.ai/ Our current focus is arc-agi 3. We recently hired the top #2 kaggler to join that effort. We are looking for people deeply passionate about AI research. I am steering our work culture towards "live-to-work": If you wouldn't be doing AI research working somewhere else, then it is not a good fit. If you don't find yourself naturally reading about AI papers or doing random AI projects in your free time, then it is not a good fit. We have high expectations and a generous equity structure to match.

Tufalabs is an independent AI research lab in Zurich, focused on high-impact academic-style AI research and large-scale LLM development. We care about compute (currently, 6× DGX B200, 2× NVL72 GB300 incoming). Our mission is to push the frontier of AI science while open-sourcing results and building the engineering muscle to pre- and post-train cutting-edge language models. The ultimate goal is automating AI research, with a strong focus on alignment and safety in the limit.

We can only sponsor visas for people with EU nationality. The position is onsite, fulltime. We are self-funded, from what i call "self-propelled machine learning": I applied ML to cool domains and made returns that way. I can give more details about this privately. If interested, email me at benjamin at tufalabs.ai

not-a-cat

vor 9 Monatendiscuss

Tufalabs | COO & Senior research/engineers | Full Time | Zurich, Onsite | 100-130k CHF | https://tufalabs.ai/

Tufalabs is an independent AI research lab in Zurich, focused on high-impact academic-style AI research and large-scale LLM development. We care about compute (currently, 6× DGX B200, 2× NVL72 GB200 incoming). Our mission is to push the frontier of AI science while open-sourcing results and building the engineering muscle to pre- and post-train cutting-edge language models. The ultimate goal is automating AI research, while keeping in mind AI alignment and safety. The ultimate goal is to make sure AI goes well for humanity.

I am looking for a COO to help scale from ~10 to ~15 people.

Ideally you care about AI research, you are excited about the idea of helping us move faster, you like to keep red-tape low, you are a technical person who can create automated processes, and have experience owning and running operations at a startup.

We are also open to senior engineers and researchers, comp negotiable

If interested, email me at benjamin at tufalabs.ai

not-a-cat

vor 9 Monatendiscuss

I founded Tufalabs about a year ago https://tufalabs.ai/ On the side, I'm making a snake game with codex-cli, and then making a ppo agent to learn to play it. Codex-cli can mostly oneshot the ppo implementation, which I find extremely impressive. Getting codex to tweak the ppo algorithm resulting in a RL agent fully solving the snake and visualizing it is very satisfying.

not-a-cat

vor 10 Monatendiscuss

Tufalabs | COO & Senior research/engineers | Full Time | Zurich, Onsite | 100-130k CHF | https://tufalabs.ai/

Tufalabs is an independent AI research lab in Zurich, focused on high-impact academic-style AI research and large-scale LLM development. We care about compute (currently, 6× DGX B200, 2× NVL72 GB200 incoming). Our mission is to push the frontier of AI science while open-sourcing results and building the engineering muscle to pre- and post-train cutting-edge language models. The ultimate goal is automating AI research, while keeping in mind AI alignment and safety. The ultimate goal is to make sure AI goes well for humanity.

I am looking for a COO to help scale from ~10 to ~15 people.

Ideally you care about AI research, you are excited about the idea of helping us move faster, you like to keep red-tape low, you are a technical person who can create automated processes, and have experience owning and running operations at a startup.

We are also open to senior researchers, comp negotiable

If interested, email me at benjamin at tufalabs.ai

not-a-cat

vor 11 Monatendiscuss

Tufalabs | COO & Senior research/engineers | Full Time | Zurich, Onsite | 100-130k CHF | https://tufalabs.ai/

Tufalabs is an independent AI research lab in Zurich, focused on high-impact academic-style AI research and large-scale LLM development. We care about compute (currently, 6× DGX B200, 2× NVL72 GB200 incoming). Our mission is to push the frontier of AI science while open-sourcing results and building the engineering muscle to pre- and post-train cutting-edge language models. The ultimate goal is automating AI research, with a strong focus on alignment and safety.

I am looking for a COO to help scale from ~10 to ~15 people.

Ideally you care about AI research, you are excited about the idea of helping us move faster, you like to keep red-tape low, you are a technical person who can create automated processes, and have experience owning and running operations at a startup.

We are also open to senior researchers, comp negotiable

If interested, email me at benjamin at tufalabs.ai

not-a-cat

letztes Jahrdiscuss

Tufalabs | AI researchers and engineers | Zurich | Full time | Onsite | $165k | https://tufalabs.ai/

We are a small, independent research group working on fundamental AI research.

Since around 2008, I always believed that AI research will be the most impactful technology to work on, both from an intellectual curiosity and technological capability perspective. After a few failed attempts to get into ML/AI research, I decided to apply it to quant finance, and then start an AI research lab from it.

This is how Tufalabs was funded about a year ago. We started working on ARC (https://arcprize.org/arc-agi), and since o3 we focus on LLM-based reasoning, self-improving loops and scaling trends. You can see some of our research there: https://tufalabs.ai/blog/_site/

We do not have too many spots left, but I want to make sure we don't miss-out on people who would be great fits. We look for obsessive people who share the same research interests. To apply, email me at not_a_cat at fastmail com.

not-a-cat

vor 2 Jahrendiscuss

Tufalabs | Research engineer | Zurich | Full Time | On-site

We are a small, independent research group working on fundamental AI research.

The plan is to first solve the ARC benchmark (https://arcprize.org/), then use that to secure more funding, and move on to more difficult benchmarks. The research topics we want to investigate next are test-time compute, active inference, LLM-based reasoning and system2 reasoning. The ultimate goal is to do cool AI research leading up to AGI.

The lab has acquired the MindsAI team recently. MindsAI is the best team in the competition so far, pioneering techniques such as fine-tuning models on synthetic data, test-time-training, and test-time-finetuning.

We are looking for people who are driven and intrinsically motivated to work on both ARC and AI/ML research. The role would include working on the main system developed by MindsAI to accelerate progress.

Compensation: $100-135k/year + generous equity plan

More info: https://tufalabs.ai/

To apply: send me an email at [email protected]

not-a-cat

vor 2 Jahrendiscuss

Hope you get it sorted. I have seen relatives struggle through it and it is a challenge

not-a-cat

vor 2 Jahrendiscuss

Unnamed AI research lab | ML research engineer | Zurich, Switzerland | Full-time | Onsite | 100-200k CHF I am starting a research lab funded by a hedge fund that I created 5 years ago. The hedge is doing well, and applies ML to quant finance. The research and trading activities should reinforce each other over time: the research publishes research and fosters in-house ML expertise. The trading applies ML to quant finance and brings funding. The research entity is incorporated, funded, has an office and 0 employees currently.

My vision for the research lab is the following:

  - A bunch of phds in a room: create a good working environment, optimize for research output & runway.
  - Take the "bitter lesson" seriously: most of the successes in modern machine learning are due to deep learning and the ability to leverage vasts amounts of compute and data.

I want to study how DL/LLM models reason, using ARC as a benchmark (https://arcprize.org/, https://arxiv.org/abs/1911.01547). Techniques I want to investigate include test time compute, active inference, meta-learning, self-play, or any technique that would be considered "system 2".

With that in mind, today I'm looking for the following role:

  - ML research engineer

Compensation will include flexible options scheme based on results. Interest and knowhow in the following is a plus: Deep learning, transformer models, reasoning, world models, learning-to-learn, mechinterp, bayesian methods, probabilistic programming.

I will mostly consider candidates having an EU passport for visa reasons.

To apply, email me at not_a_cat at fastmail dot com.

not-a-cat

vor 2 Jahrendiscuss

Unnamed AI research lab | ML research engineer | Zurich, Switzerland | Full-time | Onsite | 100-200k CHF

I am starting a research lab funded by a hedge fund that I created 5 years ago. The hedge is doing well, and applies ML to quant finance. The research and trading activities should reinforce each other over time: the research publishes research and fosters in-house ML expertise. The trading applies it to quant finance and brings funding. The research entity is incorporated, funded, has an office and 0 employees currently.

My vision for the research lab is the following:

  - A bunch of phds in a room: create a good working environment, optimize for research output & runway.
  - Take the "bitter lesson" seriously: most of the successes in modern machine learning are due to deep learning and the ability to leverage vasts amounts of compute and data.

So we make it a top priority to develop the infrastructure and expertise to be good at DL/LLMs/transformers. It doesn't mean that we don't explore new ideas. We want to learn to leverage large amounts of compute and be smart/efficient about it. If the trading works well, we can compete on an equal footing with the big labs over time in terms of amount spent in compute.

For the research itself, I want to start with the ARC challenge (https://arcprize.org/, https://arxiv.org/abs/1911.01547)

With that in mind, today I'm looking for the following role:

  - ML research engineer

Compensation will include flexible options scheme based on results.

Interest and knowhow in the following is a plus: Deep learning, transformer models, reasoning, world models, learning-to-learn, mechinterp, bayesian methods, probabilistic programming I'm also looking for evidence of past exceptional achievements.

I will mostly consider candidates having an EU passport for visa reasons.

To apply, email me at not_a_cat at fastmail dot com.

not-a-cat

vor 4 Jahrendiscuss

Discussion on twitter: https://twitter.com/shr_id/status/1602688734990053376

Submissions

Impossible Chess

The Singularity as Cognitive Decoupling

Poe-World: Compositional World Modeling with Products of Programmatic Experts

comments