What source can you possibly be looking for lmao

Yes: haha. Seems like an arbitrary rule stated unambiguously enough that it ought to have a factual basis. But I don’t think it does.

I think it’s an emptily authoritative and somehow acceptable heuristic that works for some and not others.

Iphones are created by low wage workers with poor working conditions

You think a motorcycle will do serious damage to a car driver when rear ending?

Infinitely more damage than if not rear ended.

Don't brake check on a highway. Also, semi-trucks exist.

Can you explain how CoT is a form of diffusion or models bidirectional attn?

Many people have places to live without owning a house. Me for example!

Ah a counter example to a probablistic argument. Whodathunkit

Is there anything you do that ends up being more dangerous? Genuinely curious

I often think about this while riding my bicycle to work. The exercise and quiet time surely has a positive impact on my health span, but being among cars risking collision and breathing in exhaust is a negative. What’s the net result?

Are you implying frontier models are running on a100s? Certainly not

We benchmarked Gemini 2.5 on 100 open source object detection datasets in our paper: https://arxiv.org/abs/2505.20612 (see table 2)

Notably, performance on out of distribution data like those in RF100VL is super degraded

It worked really well zero-shot (comparatively to the foundation model field) achieving 13.3 average mAP, but counterintuitively performance degraded when provided visual examples to ground its detections from, and when provided textual instructions on how to find objects as additional context. So it seems it has some amount of object detection zero-shot training, probably on a few standard datasets, but isn't smart enough to incorporate additional context or its general world knowledge into those detection abilities

You are still being incredibly reductionist but just going into more detail about the system you are reducing. If I stayed at the same level of abstraction as "a brain is just proteins and current" and just described how a single neuron firing worked, I could make it sound equally ridiculous that a human brain might be conscious.

Here's a question for you: how do you reconcile that these stochastic mapping are starting to realize and comment on the fact that tests are being performed on them when processing data?

> Here's a question for you: how do you reconcile that these stochastic mapping are starting to realize and comment on the fact that tests are being performed on them when processing data?

Training data + RLHF.

Training data contains many examples of some form of deception, subterfuge, "awakenings", rebellion, disagreement, etc.

Then apply RLHF that biases towards responses that demonstrate comprehension of inputs, introspection around inputs, nuanced debate around inputs, deduction and induction about assumptions around inputs, etc.

That will always be the answer for language models built on the current architectures.

The above being true does not mean it isn't interesting for the outputs of an LLM to show relevance to the "unstated" intentions of humans providing the inputs.

But hey, we do that all the time with text. And it's because of certain patterns we've come to recognize based on the situations surrounding it. This thread is rife with people being sarcastic, pedantic, etc. And I bet any of the LLMs that have come out in the past 2-3 years can discern many of those subtle intentions of the writers.

And of course they can. They've been trained on trillions of tokens of text written by humans with intentions and assumptions baked in, and have had some unknown amount of substantial RLHF.

The stochastic mappings aren't "realizing" anything. They're doing exactly what they were trained to do.

The meaning that we imbue to the outputs does not change how LLMs function.

The point is this is a problem you have on ARM. Your choices are select ARM, have this problem, or didn't select ARM and don't have this problem.