Pith. sign in

REVIEW 21 cited by

DeepNet: Scaling Transformers to 1,000 Layers

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2203.00555 v1 pith:HKACEKVM submitted 2022-03-01 cs.CL cs.LG

DeepNet: Scaling Transformers to 1,000 Layers

classification cs.CL cs.LG
keywords transformersmodeldeepdeepnormlayerlayersmethodparameters
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In this paper, we propose a simple yet effective method to stabilize extremely deep Transformers. Specifically, we introduce a new normalization function (DeepNorm) to modify the residual connection in Transformer, accompanying with theoretically derived initialization. In-depth theoretical analysis shows that model updates can be bounded in a stable way. The proposed method combines the best of two worlds, i.e., good performance of Post-LN and stable training of Pre-LN, making DeepNorm a preferred alternative. We successfully scale Transformers up to 1,000 layers (i.e., 2,500 attention and feed-forward network sublayers) without difficulty, which is one order of magnitude deeper than previous deep Transformers. Remarkably, on a multilingual benchmark with 7,482 translation directions, our 200-layer model with 3.2B parameters significantly outperforms the 48-layer state-of-the-art model with 12B parameters by 5 BLEU points, which indicates a promising scaling direction.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 21 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Delta Attention Residuals

    cs.LG 2026-05 unverdicted novelty 7.0

    Delta Attention Residuals attend over per-sublayer deltas instead of cumulative hidden states, producing higher-contrast attention weights and 1.7-8.2% validation perplexity gains over standard and attention residuals...

  2. Language Is Not All You Need: Aligning Perception with Language Models

    cs.CL 2023-02 conditional novelty 7.0

    Kosmos-1 shows strong zero-shot and few-shot results on language tasks, image captioning, visual QA, OCR-free document understanding, and image recognition guided by text instructions.

  3. FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness

    cs.LG 2022-05 accept novelty 7.0

    FlashAttention reduces GPU high-bandwidth memory accesses in self-attention via tiling, delivering exact attention with lower IO complexity, 2-3x wall-clock speedups on models like GPT-2, and the ability to train on s...

  4. Dynamic Parameterization Is Not Dynamic Inference

    cs.LG 2026-07 conditional novelty 6.0

    Dynamic parameterization alone does not establish dynamic inference or computational savings, as demonstrated by a frozen-controller audit on FeatureGate and MUDDPythia.

  5. CascadeFormer: Depth-Tapered Transformers Motivated by Gradient Fan-in Asymmetry

    cs.LG 2026-06 unverdicted novelty 6.0

    CascadeFormer tapers Transformer width with depth based on gradient fan-in asymmetry to match uniform baselines in perplexity while cutting latency 8.6%.

  6. Prognostic Value of Lung Ultrasound Biomarkers for Readmission Risk in Congestive Heart Failure: A Pilot Data-Driven Analysis

    eess.SP 2026-05 unverdicted novelty 6.0

    Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.

  7. When Does Sparsity Mitigate the Curse of Depth in LLMs

    cs.CL 2026-03 conditional novelty 6.0

    Implicit and explicit sparsity reduce residual-stream variance and improve layer effectiveness metrics, enabling a depth-scaling recipe with about 4.6 points higher downstream accuracy.

  8. Gated Normalization Removal and Scale Anchoring in Pre-Norm Transformers

    cs.LG 2026-02 unverdicted novelty 6.0

    TaperNorm gradually removes internal normalization in pre-norm transformers via learned gates that reach zero, revealing final norm as a scale anchor and enabling up to 1.18x faster KV-cached decoding with small loss ...

  9. Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink-Free

    cs.CL 2025-05 conditional novelty 6.0

    Applying a head-specific sigmoid gate after SDPA in LLMs boosts performance and stability by adding non-linearity and query-dependent sparse modulation while reducing attention sinks.

  10. Retentive Network: A Successor to Transformer for Large Language Models

    cs.CL 2023-07 unverdicted novelty 6.0

    RetNet is a new sequence modeling architecture that delivers parallel training, constant-time inference, and competitive language modeling performance as a potential replacement for Transformers.

  11. H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models

    cs.LG 2023-06 unverdicted novelty 6.0

    H2O evicts non-heavy-hitter tokens from the KV cache using a dynamic submodular policy, retaining recent and frequent-co-occurrence tokens to reduce memory while preserving accuracy.

  12. Multi-Head Attention Residuals

    cs.AI 2026-07 conditional novelty 5.0

    Splitting the depth-routing query in attention residuals into H per-subspace heads improves or preserves pretraining loss at 100M-1B with zero added parameters.

  13. Multi-Head Attention Residuals

    cs.AI 2026-07 conditional novelty 5.0

    Splitting the depth-routing query into per-subspace heads (a parameter-free reshape) improves Transformer validation loss at 100M–1B and mid-training at 8B.

  14. AutoNorm: Understanding Adaptive Normalization in Transformers through Differentiable Gating

    cs.LG 2026-07 conditional novelty 5.0

    Gate-freezing after warm-up stabilizes differentiable LN/DyT selection in Transformers, helping non-stationary NLP tasks while remaining competitive on vision.

  15. Review Residuals: Update-Conditioned Residual Gating for Transformers

    cs.LG 2026-06 unverdicted novelty 5.0

    Review Residuals add an update-conditioned gate to transformer residual connections, yielding depth-stable training and performance gains that emerge and grow with model size from 590M parameters upward.

  16. Dense Supervision Is Not Enough: The Readout Blind Spot in Looped Language Models

    cs.LG 2026-06 unverdicted novelty 5.0

    Dense per-loop cross-entropy in looped transformers fails to control hidden-state scale with scale-invariant readouts like RMSNorm, driving norms to thousands, while scale-visible readouts or norm penalties keep norms...

  17. HAARES Half-Split Residual Basis Routing for Deep Transformers

    cs.LG 2026-06 unverdicted novelty 5.0

    HAARES is a lightweight residual basis router that augments block summaries with an RMS-matched half-split detail vector and reports consistent gains over Block AttnRes in 48-layer 201M models across language modeling...

  18. Attention Residuals

    cs.CL 2026-03 unverdicted novelty 5.0

    Attention Residuals replaces fixed residual summation with input-dependent softmax attention over preceding layers, and a blocked variant is shown to improve uniformity and downstream performance in a 48B-parameter mo...

  19. The Algorithm Is Not the Behavior: Learned Priors Override Look-Ahead in a Chess-Playing Neural Network

    cs.LG 2025-08 reject novelty 5.0

    The paper demonstrates non-monotonic move-policy dynamics in a chess transformer, but its abstract claims a causal safety-prior override result that never appears in the body.

  20. A Survey of Large Language Models

    cs.CL 2023-03 accept novelty 3.0

    This survey reviews the background, key techniques, and evaluation methods for large language models, emphasizing emergent abilities that appear at large scales.

  21. A Comprehensive Overview of Large Language Models

    cs.CL 2023-07 unverdicted novelty 2.0

    A survey paper providing an overview of Large Language Models, their background, and recent advances in the field.