Hacker News

Latent Lattice Language Models

56 points by E-Reverance ago | 5 comments

augment_me |next [-]

Seems like just an extension of Short-conv in the linear attention methods where we instead carry the EMA of the entire sequence instead of nearest 4 tokens.

It would make sense in my opinion to do the evals based on the reason for needing Short-conv in the architectures that use it in the first place, where it was added to support exact retrieval/needle-in-haystack when we compress the attention state.

It's odd to make this arch extension and not test beyond a 2k context window, or with the needle-in-haystack benchmark.

In my head this should fail because the EMA cannot carry this granular information

royal__ |next |previous [-]

How significant is a .98 decrease in perplexity. I'm a researcher in this field and I'm having trouble interpreting the implication of this.

tkslucas |next |previous [-]

Interesting approach to scale compute, never thought of using the per layer state like that

amitport |next |previous [-]

Isn't that "Infini-attention"? https://arxiv.org/abs/2404.07143 (2024)

bionhoward |previous [-]

I am so excited for RNNs to be cool again