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

Disagreement among frontier LLMs on real-world fact-checks

lenz.io

Post reply on HN
Print view View original

21–30 of 377 posts

Re: Disagreement among frontier LLMs on real-world fact-checks

#21

Post by embedding-shape »

kostaj wrote: post #20
embedding-shape wrote: post #9

> These aren't benchmark items with public answer keys — they're claims real users submitted for verification to a fact-checking platform. Cool. I wonder if anything of this matters when the authors don't disclose exactly how much of their report was written and made with LLMs in the first place? There even is a "11. Ethics & data use" section, and the research is about LLMs being infallible in some ways, yet the usa…

Data collection and processing was done manually. LLMs helped with the report drafting. Everything was human reviewed before publishing.

So it's not a secret, why you don't add this upfront to the report? The report itself is even about LLMs, makes a lot of sense to disclose your usage of them for writing the report, especially when you're presenting evidence that boils down to LLMs being infallible.

Re: Disagreement among frontier LLMs on real-world fact-checks

#22

Post by jiggawatts »

kostaj wrote: post #2

Author here. 67% (95% CI 64–70%) of 1,000 recent real user claims to a fact-checking platform had at least one of GPT-5.4, Claude Opus 4.7, Gemini 3 Pro, Gemini 3 Pro+Search, and Sonar Pro dissent from the panel majority — or no majority formed at all. Panel-level Krippendorff's α (ordinal) = 0.639, i.e. nontrivial but limited agreement. Quick context on what's in the writeup and what isn't: - What's measured: parsed…

Many of the rows in that spreadsheet reference "current events", which models aren't expected to do much better at than a human making an educated guess! They all have cutoff dates either last year or early this year and know nothing about what happened in "April 2026".

This is doubly problematic because you evaluated earlier models like Gemini Pro 3 instead of 3.1, GPT 5.4 instead of 5.5, etc...

Given that it's only a thousand short questions, you should be able to re-run your test in about an hour with the latest models, so... why haven't you?

Similarly, LLM output is non-deterministic, so if you could get more interesting stats of your data set by repeating each question 'n' times for each model.

Re: Disagreement among frontier LLMs on real-world fact-checks

#23

Post by harpastum »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

Without providing definitions of "True / Mostly True / Misleading / False" to each rater, I rate the article's claim that "Only one verdict bucket can be correct per claim" as false.

Something can be simultaneously "misleading" and either true or false. Which category should something go in if it's "mostly false"?

How much can something be wrong before it goes from "mostly true" to "false" (objectively, both have some part of the fact that is not true)?

This is at least partly testing the model's definition of "mostly" and "misleading". Not its understanding of the fact. Claiming that this means the models have fundamental disagreement on the facts themselves is an overreach.

Re: Disagreement among frontier LLMs on real-world fact-checks

#24

Post by tosh »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

ty for digging this up, appreciate the time saving

Re: Disagreement among frontier LLMs on real-world fact-checks

#25

Post by singpolyma3 »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

False vs misleading doesn't seem like a disagreement?

Re: Disagreement among frontier LLMs on real-world fact-checks

#26

Post by embedding-shape »

harpastum wrote: post #23
simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

Without providing definitions of "True / Mostly True / Misleading / False" to each rater, I rate the article's claim that "Only one verdict bucket can be correct per claim" as false . Something can be simultaneously "misleading" and either true or false. Which category should something go in if it's "mostly false"? How much can something be wrong before it goes from "mostly true" to "false" (objectively, both have so…

> I guess the goal is to test the models and not the harness

Less important than the harness, is the system/user prompts themselves (which of course, are put in the harness), which is effectively what this study seems to be testing. With a better prompt, I'm sure the models would look more the same to each other, as the biggest/best models have more or less identical strong prompt-adherence in my experience.

Re: Disagreement among frontier LLMs on real-world fact-checks

#27

Post by kostaj »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

Used "No explanations, no qualifiers." to force the models to answer only with one of the four labels. It's worth running a separate test with more explanation in the prompt on how to classify between the four buckets.

Re: Disagreement among frontier LLMs on real-world fact-checks

#29

Post by malfist »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

> All almonds are grown in the U.S. state of California

This isn't misleading, it's flat out false. Characterizing misleading as also acceptable isn't valid here. If you go an ask anyone on the street if this is true, false or misleading, I'm sure almost everyone would say it's false. After all, I can grow almonds myself.

Re: Disagreement among frontier LLMs on real-world fact-checks

#30

Post by jannyfer »

simonw wrote: post #11

Here's the prompt they used: Classify this claim as of : " " Output exactly one label: True, Mostly True, Misleading, or False. No explanations, no qualifiers. The claims look like this: https://lenz.io/research/llm-disagreement/data.csv I put that in Datasette Lite to make it easier to explore. Here's an example of a disagreement: https://lite.datasette.io/?csv=https%3A%2F%2Fstatic.simonwil... The claim was "All alm…

Thank you, my eyes glazed over when I saw the article was written with AI.
Post reply on HN