AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Hungry Hungry Hippos: Towards Language Modeling with State Space Models

Daniel Y Fu, Tri Dao, Khaled Kamal Saab, Armin W Thomas, Atri Rudra, Christopher Re

ICLR 2023 · Architectures and efficiency › Attention and transformers · approach: State space and recurrent sequence models · 119 citations (OpenAlex)

This paper tried a common change: Add attention module

Add two attention layers to H3 (hybrid H3-attention) · OpenWebText · perplexity (lower better): 20.6 to 19.6 vs Transformer Improved

What 492 accepted papers report for this change

See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add two attention layers to H3 (hybrid H3-attention)H3 125M · OpenWebTextperplexity (lower better): 20.6 to 19.6 vs TransformerImproved
Use H3 layer instead of S4D/GSS SSM layers125M language model · OpenWebTextperplexity (lower better): 21.0 vs 24.9 (S4D) and 24.0 (GSS)Improved
Pure H3 vs Transformer at 125MH3 125M · OpenWebTextperplexity (lower better): 21.0 vs 20.6 (within 0.4)No clear effect
Add H3 layer for synthetic associative recall2-layer models · Associative Recall syntheticaccuracy: 99.8 vs 86.0 (S4D) and 78.0 (GSS)Improved
Add H3 layer for synthetic induction head2-layer models · Induction Head syntheticaccuracy: 100.0 vs 35.6 (S4D) and 6.8 (GSS)Improved
FlashConv (fused block FFT + state passing) applied to S4S4 · Long Range Arenaspeedup over Transformer: 2.9× to 5.8×Improved
Hybrid H3-attention instead of Transformer at 1.3B inference1.3B language model · Inference throughput, prompt length 1536tokens/s (higher better): 1240 vs 520 (about 2.4×)Improved
Hybrid H3-attention vs Transformer at 2.7B on Pile scale2.7B language model · Pile perplexityperplexity (lower better): 5.4 vs 5.7 (GPT-Neo-2.7B)Improved
Hybrid H3-attention vs Transformer on SuperGLUE zero-shot average at 2.7B2.7B language model · SuperGLUE zero-shotaverage accuracy (higher better): 56.8 vs 55.5 (OPT-2.7B) and 54.6 (GPT-Neo-2.7B)Improved
Hybrid H3 with 2 attention layers vs pure H3 at 125M (ablation)125M language model · OpenWebTextperplexity (lower better): 19.6 vs 21.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy