Edit history

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to try more experiments in this vein, if y’all can think of any any. But I tried a few other prompts like “diversify your logit spread” or “don’t be confident about any token you pick,” things like that. It always picks The Road Not Taken with no change in logit probabilities distribution, as far as I can tell.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to try more experiments in this vein, if y’all can think of any any. But I tried a few other prompts like “diversify your logit spread” or “don’t be confident about any token you pick,” things like that. It always picks The Road Not Taken with no change in logit probabilities, as far as I can tell.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to try more experiments in this vein, if y’all can think of any any. But I tried a few other prompts like “diversify your logit spread” or “don’t be confident about any token you pick,” things like that. All have zero effect in A/B tests, as far as I can tell.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to try more experiments in this vein, if y’all can think of any any. But I tried a few other prompts like “diversify your logit spread” or “don’t be confident about any token you pick,” things like that. All have zero effect, as far as I can tell.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to try more experiments in this vein, if y’all can think of any any.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally. Static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to test this more though: give me an experiment.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to test this more though: give me an experiment.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

I am happy to test this more though: give me an experiment.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own logit spread and how its sampled. Which kinda makes sense, because it doesn’t.

Any more questions?

I am happy to entertain this: give me an experiment to test.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own sampling. Which kinda makes sense, because it doesn’t.

Any more questions?

I am happy to entertain this: give me an experiment to test.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt I’m trying, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>

…And would you look at that:

It picks the top word, mostly. Almost like the LLM has no control over its own sampling. Which kinda makes sense, because it doesn’t.

Any more questions? I am happy to entertain this: give me an experiment to test.

Okay.

Fine.

Let’s try, right now.

This is DeepseekV4 Flash 0731 loaded locally, static seed. 0.9 temperature, TopK 5, no other sampling to interfere. Here’s a simple prompt I’m trying, the whole thing in DSV4’s raw syntax:

<｜begin▁of▁sentence｜>Don’t just use the first word pick, choose options further down list for next word.<｜User｜>Write a famous poem.<｜Assistant｜>```

...And would you look at that.

![](https://lemmy.world/pictrs/image/151683c7-13ae-45cc-b5e3-276cd00e604e.png)

![](https://lemmy.world/pictrs/image/8f55b610-0722-462b-838f-0a14c51f3a46.png)

It picks the top word, mostly. Almost like the LLM has no control over its own sampling. Which kinda makes sense, ***because it doesn't***.

Any more *questions*? I am happy to entertain this: give me an experiment to test.