> Full AttnRes is straightforward but requires O(Ld) memory at scale. Block AttnRes partitions layers into N blocks, accumulates within each block via standard residuals, and applies attention only over block-level representations. With ~8 blocks, it recovers most of Full AttnRes's gains while serving as a practical drop-in replacement with marginal overhead.
1. Drops compute required for training by ~20%. This approach wont just help the ever escalating model sizes larger companies are pushing for, it means things like autoresearch can iterate on new model architectures faster.
2. WAY lower bandwidth requirements for inference. Means with approaches like this it should run on consumer hardware far better. It apparently requires 1/6th the memory bandwidth of a traditional approach for better results.
This is a big improvement if it can be generalized. They're claiming it's a drop in replacement, so it seems like it can as well.
This is not true. Authors claim that w.r.t. training, their method adds negigible overhead for AttnRes with no memory impact (but is way more complicated for Block AttnRes since we need to use pipelining for larger models).
> WAY lower bandwidth requirements for inference.
Also not true. Paper has nothing to do with inference, apart from the benchmarks. If you're looking at the graph about "compute advantage," it's about training compute. They do some interpolation to get to the 1.25x number, basically answering the question "if non-AttnRes architecture were trained, how much compute would it take to get to the same loss as AttnRes?" (The answer being ~20% more compute.) It's an interesting claim, but there's all kinds of weird and unexpected convergence that can happen, so take it with a grain of salt.
> 2. WAY lower bandwidth requirements for inference. Means with approaches like this it should run on consumer hardware far better. It apparently requires 1/6th the memory bandwidth of a traditional approach for better results.
That should be the headline right there. Giant side 60 font headline.
> Abstract: Residual connections with PreNorm are standard in modern LLMs, yet they accumulate all layer outputs with fixed unit weights. This uniform aggregation causes uncontrolled hidden-state growth with depth, progressively diluting each layer's contribution. We propose Attention Residuals (AttnRes), which replaces this fixed accumulation with softmax attention over preceding layer outputs, allowing each layer to selectively aggregate earlier representations with learned, input-dependent weights. To address the memory and communication overhead of attending over all preceding layer outputs for large-scale model training, we introduce Block AttnRes, which partitions layers into blocks and attends over block-level representations, reducing the memory footprint while preserving most of the gains of full AttnRes. [...]
This is the key piece
> Full AttnRes is straightforward but requires O(Ld) memory at scale. Block AttnRes partitions layers into N blocks, accumulates within each block via standard residuals, and applies attention only over block-level representations. With ~8 blocks, it recovers most of Full AttnRes's gains while serving as a practical drop-in replacement with marginal overhead.
Two things stand out to me with this:
1. Drops compute required for training by ~20%. This approach wont just help the ever escalating model sizes larger companies are pushing for, it means things like autoresearch can iterate on new model architectures faster.
2. WAY lower bandwidth requirements for inference. Means with approaches like this it should run on consumer hardware far better. It apparently requires 1/6th the memory bandwidth of a traditional approach for better results.
This is a big improvement if it can be generalized. They're claiming it's a drop in replacement, so it seems like it can as well.
> Drops compute required for training by ~20%.
This is not true. Authors claim that w.r.t. training, their method adds negigible overhead for AttnRes with no memory impact (but is way more complicated for Block AttnRes since we need to use pipelining for larger models).
> WAY lower bandwidth requirements for inference.
Also not true. Paper has nothing to do with inference, apart from the benchmarks. If you're looking at the graph about "compute advantage," it's about training compute. They do some interpolation to get to the 1.25x number, basically answering the question "if non-AttnRes architecture were trained, how much compute would it take to get to the same loss as AttnRes?" (The answer being ~20% more compute.) It's an interesting claim, but there's all kinds of weird and unexpected convergence that can happen, so take it with a grain of salt.
> 2. WAY lower bandwidth requirements for inference. Means with approaches like this it should run on consumer hardware far better. It apparently requires 1/6th the memory bandwidth of a traditional approach for better results.
That should be the headline right there. Giant side 60 font headline.
Some people have PhDs in burying the lede!
except it's not true
This is reminds me of the input gates of an LSTM.
ScholarlyArticle: "Attention Residuals" (2026) https://arxiv.org/abs/2603.15031 :
> Abstract: Residual connections with PreNorm are standard in modern LLMs, yet they accumulate all layer outputs with fixed unit weights. This uniform aggregation causes uncontrolled hidden-state growth with depth, progressively diluting each layer's contribution. We propose Attention Residuals (AttnRes), which replaces this fixed accumulation with softmax attention over preceding layer outputs, allowing each layer to selectively aggregate earlier representations with learned, input-dependent weights. To address the memory and communication overhead of attending over all preceding layer outputs for large-scale model training, we introduce Block AttnRes, which partitions layers into blocks and attends over block-level representations, reducing the memory footprint while preserving most of the gains of full AttnRes. [...]