Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Handbook.md shows that long policy documents do not reliably govern agents

arxiv.org

Post reply on HN
Print view View original

171–180 of 237 posts

Re: Handbook.md shows that long policy documents do not reliably govern agents

#171

Post by epolanski »

wongarsu wrote: post #26

Any model with a good score on this benchmark would have a good claim on superhuman abilities. Humans are pretty terrible at being thrown a long policy document and being expected to follow it And while we shouldn't anthropomorphize these models too much, I wouldn't be surprised if many of the core reasons for failures are similar. Working memory is a limited resource; you can only focus on so many things at once; re…

> Any model with a good score on this benchmark would have a good claim on superhuman abilities. Humans are pretty terrible at being thrown a long policy document and being expected to follow it

Hmmm, if AI has to grow in the workplace, it has to follow processes to the letter, yet claude code forgets by the second turn my "don't commit" prompt.

Of course claude code is a generic and crap harness over terrific models that can't fit bureaucratic processes, and is increasingly worse at doing so anyway since Opus 4.6 peak.

Re: Handbook.md shows that long policy documents do not reliably govern agents

#173

Post by Aurornis »

ux266478 wrote: post #150
Aurornis wrote: post #118

Earlier quoted context omitted.

> Well the answer is that VC-backed companies Look, I enjoy local LLMs as much as anyone, but I think there is some motivated reasoning happening in this thread to try to make local LLMs sound like a utopia against those evil VCs. Local LLMs suffer from the same problems.

And I think you didn't understand what I wrote, so let me reiterate: Of course policies are equally ineffective at strictly governing the behavior of local models. Why would that change? The actual problem is that somebody is attempting to misuse them for that in the first place. There are only two reasons for it: - They have no other options OR - They have no idea what they're doing Local models solve the first prob…

> In-context, your reply reads like you're asserting that anthropic's offerings give you all the same control as a local model, and that having this baseline of control over a computer program is "utopian".

No, I’m talking about the idea up thread that having control over more settings removes all LLM defects with local models.

I have all the control I want over the local models they run. No combination of settings is going to make them outperform Anthropic’s models.

Re: Handbook.md shows that long policy documents do not reliably govern agents

#174

Post by Aurornis »

nzeid wrote: post #158
Aurornis wrote: post #115

Earlier quoted context omitted.

> Want it to go away, almost like magic? Local inference. When its under your control, and no longer being forced to hold it wrong, all of the common LLM defects will go away. This is just not true. Any local LLM you can host on consumer-accessible hardware has all of these defects, too. Adjusting the knobs doesn’t solve everything. The closest you can get to frontier performance is Kimi K3, but you’re not hosting th…

This is a bit of a strawman. The parent comment wasn't suggesting that the inherent defects magically disappear nor was it suggesting that local inference today is sufficient for all tasks. At the heart of it, self-hosting liberates your use cases from all the horribly opaque configuration, shadow prompting, etc. And local models are only getting better and more diverse every month.

> This is a bit of a strawman. The parent comment wasn't suggesting that the inherent defects magically disappear

The parent comment literally said that the common LLM defects would go away.

Direct quote:

> all of the common LLM defects will go away

Re: Handbook.md shows that long policy documents do not reliably govern agents

#175

Post by KerrAvon »

Aurornis wrote: post #115
DiabloD3 wrote: post #6

This is a problem with long context models. To put it as simple and as bluntly as possible: just because they claim you can use 1M tokens in your context doesn't mean its true and you should do that. Due to extreme quantization of models and the context's KV cache, and also just really shitty samplers provided to the user (hell, most are just getting rid of sampler knobs altogether), this problem will absolutely cont…

> Want it to go away, almost like magic? Local inference. When its under your control, and no longer being forced to hold it wrong, all of the common LLM defects will go away. This is just not true. Any local LLM you can host on consumer-accessible hardware has all of these defects, too. Adjusting the knobs doesn’t solve everything. The closest you can get to frontier performance is Kimi K3, but you’re not hosting th…

We're on the cusp of Kimi K3 becoming usable on sub-10k hardware.

https://github.com/gavamedia/deltafin

Re: Handbook.md shows that long policy documents do not reliably govern agents

#176

Post by edot »

KerrAvon wrote: post #175
Aurornis wrote: post #115

Earlier quoted context omitted.

> Want it to go away, almost like magic? Local inference. When its under your control, and no longer being forced to hold it wrong, all of the common LLM defects will go away. This is just not true. Any local LLM you can host on consumer-accessible hardware has all of these defects, too. Adjusting the knobs doesn’t solve everything. The closest you can get to frontier performance is Kimi K3, but you’re not hosting th…

We're on the cusp of Kimi K3 becoming usable on sub-10k hardware. https://github.com/gavamedia/deltafin

14 seconds per token? Not tokens per second. Seconds per token? That’s nowhere near the cusp!

Re: Handbook.md shows that long policy documents do not reliably govern agents

#177

Post by ekidd »

KerrAvon wrote: post #175
Aurornis wrote: post #115

Earlier quoted context omitted.

> Want it to go away, almost like magic? Local inference. When its under your control, and no longer being forced to hold it wrong, all of the common LLM defects will go away. This is just not true. Any local LLM you can host on consumer-accessible hardware has all of these defects, too. Adjusting the knobs doesn’t solve everything. The closest you can get to frontier performance is Kimi K3, but you’re not hosting th…

We're on the cusp of Kimi K3 becoming usable on sub-10k hardware. https://github.com/gavamedia/deltafin

For values of "usable" that include "14.6 seconds/token". It's a cool accomplishment! And newer hardware would speed it up some. But I think I'd want something a bit faster before declaring it usable in practice.

Re: Handbook.md shows that long policy documents do not reliably govern agents

#178

Post by eikenberry »

DiabloD3 wrote: post #6

This is a problem with long context models. To put it as simple and as bluntly as possible: just because they claim you can use 1M tokens in your context doesn't mean its true and you should do that. Due to extreme quantization of models and the context's KV cache, and also just really shitty samplers provided to the user (hell, most are just getting rid of sampler knobs altogether), this problem will absolutely cont…

It seems like you mean to use your own agent (where you control the prompts), not local inference. A custom agent works just as well with cloud based models.

Re: Handbook.md shows that long policy documents do not reliably govern agents

#179

Post by drob518 »

trallnag wrote: post #145
msejas wrote: post #64

Earlier quoted context omitted.

Most people jump straight to agents when what they actually need is a graph. Example: a mining company receives free-text reports from field geologists. You could have: Geologist report -> LLM call extracts minerals we are looking for (you inject a db query result on the user prompt), locations, assay mentions and risks into structured fields -> LLM call classifies evidence into positive indicators, negative indicato…

Can I have graphs as the default and fall back to agentic behavior if the prompt / task can't be mapped to a known process graph?

Sure. You could easily have the model choose a workflow/graph for execution and then call a tool that implements the execution engine for that. The tool then makes calls for model inferencing at various points in the processing.

Re: Handbook.md shows that long policy documents do not reliably govern agents

#180

Post by edot »

msejas wrote: post #64
drob518 wrote: post #43

Earlier quoted context omitted.

What do you mean by a graph of one shot prompts?

Most people jump straight to agents when what they actually need is a graph. Example: a mining company receives free-text reports from field geologists. You could have: Geologist report -> LLM call extracts minerals we are looking for (you inject a db query result on the user prompt), locations, assay mentions and risks into structured fields -> LLM call classifies evidence into positive indicators, negative indicato…

Isn’t this just software? But with an LLM at various points instead of a deterministic call to some function?
Post reply on HN