[bt1a]

This won't dethrone Llama 3, but it's equally impressive.

They mention this model's relative weakness in the TruthfulQA eval, since it's more lossy trying to pack 'knowledge' into a small model relative to problem-solving skills (which shine on MMLU)

Regardless - still a very useful thing to have offline and on the fly. Those scores are nothing to scoff at.

Given that these pipelines are likely harder harder to imitate than new architectures like Transformers, I assume there has been and will be an intense focus on synthetic data generation and cleansing. Llama 3 used 15T of tokens in its training corpus vs 4.8T in the "scaled-up" version of phi-3. If you made it to the end of this disjointed ramble I'm sorry

[Grimblewald]

Even llama3 has its issues. Ive been quite impressed so far but if the context gets a little long it freaks out, gets stuck repeating the same token or just fails to finish an answer. This is for the full f16 8B model, so it cant be put down to quantization. It also doesnt quite handle complex instructions as well as the benchmarks would imply should.

[andai]

Supposedly LLMs (especially smaller ones) are best suited to tasks where the answer is in the text, i.e. summarization, translation, and answering questions.

Asking it to answer questions on its own is much more prone to hallucination.

To that end I've been using Llama 3 for summarizing transcripts of YouTube videos. It does a decent job, but... every single time (literally 100% of the time), it will hallucinate a random name for the speaker.* Every time! I thought it might be the system prompt, but there isn't one.

My own prompt is just "{text}\n\n###\n\nPlease summarize the text above."

If I ask it to summarize in bullet points, it doesn't do that.

I'm assuming there was something in the (instruct) training data that strongly encourages that, i.e. a format of summaries beginning with the author's name? Seems sensible enough, but obviously backfires when there's literally no data and it just makes something up...

*In videos where the speaker's name isn't in the transcript. If it's a popular field, it will often come up with something plausible (e.g. Andrew Ng for an AI talk.) If it's something more obscure, it'll dream up something completely random.

[kiratp]

The technique to use is to give the model an “out” for the missing/negative case.

"{text}\n\n###\n\nPlease summarize the text above. The text is a video transcript. It may not have the names of the speakers in it. If you need to refer to an unnamed speaker, call them Speaker_1, Speaker_2 and so on."

[woodson]

Especially for small models I had very bad results for use in translation. Even trying all kinds of tricks didn’t help (apparently prompting in the target language helps for some). Encoder-decoder models such as FLAN-T5 or MADLAD-400 seemed far superior at equal or even smaller model size.

[andai]

I forget which model (LLaMA 3?) but I heard 95% of the training data was English.

[Grimblewald]

for sure, so my use case for example is

"using the following documentation to guide you {api documentation}, edit this code {relevant code}, with the following objective: Replace uses of {old API calls} in {some function} with with relevant functions from the supplied documentation"

It mostly works, but if the context is a little to long, sometimes it will just spam the same umlaut or number (always umlaut's or numbers) over and over for example. Perhaps some fine-tuning of parameters like temp. or repetition penalty might fix it, time will tell.

[andai]

Are you using ollama? Devs said there was a bug that occurs when context is full, they're working on it.

[Grimblewald]

That would do it, I am indeed.

[behohippy]

I had this same issue with incomplete answers on longer summarization tasks. If you ask it to "go on" it will produce a better completion, but I haven't seen this behaviour in any other model.

[Grimblewald]

Neither, still, the answers it does provide - despite a few hiccups - are truly outstanding. I am really impressed with this model, even with its issues. Though, I am sure the issues such that they are, are a month or two away from being fixed. For what its worth, I haven't played as much with the bigger model, but it seems to not struggle with the same, though take that with a grain of salt, it runs too slow on my hardware for me to rapidly test things.

[IvanAchlaqullah]

> TruthfulQA

Wait, people still use this benchmark? I hear there's a huge flaw on it.

For examples, fine-tuning the model on 4chan make it scores better on TruthfulQA. It becomes very offensive afterwards though, for obvious reasons. See GPT-4chan [1]

[1] https://www.youtube.com/watch?v=efPrtcLdcdM

[thomashop]

Couldn't it be that training it on 4chan makes it more truthful for some reason?

[wongarsu]

Could it be that people who can talk anonymously with no reputation to gain or lose and no repercussions to fear actually score high on truthfulness? Could it be that truthfulness is actually completely unrelated to the offensiveness of the language used to signal in-group status?

[cptcobalt]

This unironically feels like good research & paper potential.

[andy99]

Not sure I understand your example? It's not an offensiveness benchmark, in fact I can imagine a model trained to be inoffensive would do worse on a truth benchmark. I wouldn't go so far as to say truthfulQA is actually testing how truthful a model is or its reasoning. But it's one of the least correlated with other benchmarks which makes it one of the most interesting. Much more so than running most other tests that are highly correlated with MMLU performance. https://twitter.com/gblazex/status/1746295870792847562

[nurumaik]

>scores better

>very offensive

Any cons?

[hoseja]

Looks like a good and useful benchmark.

[andai]

"Omit that training data..."