Attention Sinks and Outliers in Attention Residuals

Attention Residuals replace standard additive residual connections with learned softmax attention over previous layer outputs, enabling selective cross-layer routing. However, standard Attention Residuals still attend over cumulative hidden states in previous layers, which are highly redundant. We show that this redundancy leads to routing collapse in deeper layers: attention weights become low-contrast and closer to uniform (max weight ${\approx}$0.2), limiting the model's ability to select informative states in previous layers. This raises a key but underexplored design question: what layer-wise representations should be routed in Attention Residuals? To answer this question, we propose Delta Attention Residuals, which attend over deltas -- the change introduced by each sublayer ($\mathbf{v}_i = \mathbf{h}_{i+1} - \mathbf{h}_i$) -- instead of cumulative states. Delta representations are structurally diverse and yield higher-contrast attention distributions (max weight ${\approx}$0.6), enabling more selective and effective routing across layers. This principle applies at both per-sublayer and block granularity. Across all tested scales (220M--7.6B), Delta Attention Residuals consistently outperform both standard residuals and Attention Residuals, with 1.7--8.2% validation perplexity gains. Delta Attention Residuals also enables converting pretrained checkpoints into Delta Attention Residuals via standard fine-tuning. Code is available at https://github.com/wdlctc/delta-attention-residuals-code.

Attention Residuals (AttnRes) replace the fixed residual sum with depth-wise attention over previous sub-layer outputs in Large Language Models (LLMs), but use each output as both a full-dimensional key and value. This couples routing with representation and makes the cost of computing depth-routing scores scale with hidden width $d$. We propose Low-Rank Attention Residuals (LR-AttnRes), which keep full-dimensional residual values while using $r$-dimensional keys, with $r < d$, for routing. LR-AttnRes uses the last $r$ dimensions of each value as the routing key, reducing total residual-side FLOPs while still improving performance. Comprehensive sweeps across the number of blocks ($N$) and $r$ show that depth-wise routing can be effective with far fewer dimensions than the model width. At both $1$B and $4$B parameters with $r = d/4$, LR-AttnRes achieves lower final validation loss, higher average downstream accuracy, and higher measured training-step throughput than standard AttnRes. We also provide a fused kernel supporting standard and low-rank routing. We release all code, the kernel, and all trained models to facilitate future research.

When attention concentrates on a single token, a sink, what is the model actually computing? Attention sinks are ubiquitous in softmax transformers, yet this shared visual signature can hide fundamentally different algorithms. We show that visually similar sink patterns can reflect two distinct mechanisms: {i} adaptive nop, where a head suppresses its update by routing to a null token, and {ii} broadcast, where a sink aggregates and redistributes global information. In that case, sinks serve an analogous role: a safe destination when there is nothing useful to compute. Proposed interventions like gating or registers work because they implicitly target one or the other, revealing a duality between method and assumed mechanism: gating implicitly assumes nop; registers implicitly assume broadcast. Each mechanism leaves distinct traces (nop sinks exhibit negligible value norms; broadcast sinks induce low-rank outputs) which we formalize on synthetic tasks and use to derive practical diagnostics. Applied to pretrained vision transformers, these diagnostics reveal that both mechanisms exist at scale: sinks transition from CLS in early layers to patches in deeper layers, and concentrate in specialized heads. Strikingly, register tokens, designed for broadcast, are repurposed to also serve nop, confirming that neither intervention alone suffices. Combining gating with registers yields complementary gains in stability and performance. Overall, we find that the same attention pattern can reflect two very different computations and effective intervention requires first asking what the model is actually computing.