Pith. sign in

REVIEW 22 cited by

Query-Key Normalization for Transformers

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 2010.04245 v1 pith:XONKRPS7 submitted 2020-10-08 cs.CL cs.AIcs.LG

classification cs.CLcs.AIcs.LG
keywords normalizationdimensionlow-resourcetranslationadaptingalongapplyarbitrary
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
abstract

Low-resource language translation is a challenging but socially valuable NLP task. Building on recent work adapting the Transformer's normalization to this setting, we propose QKNorm, a normalization technique that modifies the attention mechanism to make the softmax function less prone to arbitrary saturation without sacrificing expressivity. Specifically, we apply $\ell_2$ normalization along the head dimension of each query and key matrix prior to multiplying them and then scale up by a learnable parameter instead of dividing by the square root of the embedding dimension. We show improvements averaging 0.928 BLEU over state-of-the-art bilingual benchmarks for 5 low-resource translation pairs from the TED Talks corpus and IWSLT'15.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 22 Pith papers

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

  1. One QK Channel, Many Sources: Guarding Low-Precision Attention Collapse

    cs.LG 2026-08 conditional novelty 7.0 of 10

    Different low-precision errors converge on the same query-key spectral runaway, entry is gated by temporal sign-coherence, and a dormant query-key normalization guard contains it.

  2. Robust Filter Attention: Self-Attention as Precision-Weighted State Estimation

    cs.LG 2025-09 unverdicted novelty 7.0 of 10

    Robust Filter Attention models self-attention as consistency-based state estimation under a linear SDE for token trajectories, matching standard attention complexity while showing lower perplexity and better zero-shot...

  3. Characterization and Mitigation of Training Instabilities in Microscaling Formats

    cs.LG 2025-06 conditional novelty 7.0 of 10

    Block-scaled MX low-precision training is unstable because quantization of tightly clustered layer-norm weights and some activations injects multiplicative gradient bias, and this can be fixed by keeping activations i...

  4. Absolute Coordinates Make Motion Generation Easy

    cs.CV 2025-05 conditional novelty 7.0 of 10

    Using absolute 3D joint coordinates with a plain Transformer and velocity-prediction diffusion outperforms the standard local-relative motion representation, improving fidelity and enabling direct control.

  5. Sona Technical Report

    cs.IR 2026-08 conditional novelty 6.0 of 10

    Sona, a single model that both generates and ranks recommendations from one user representation, beat a production cascade of more than 15 components in a live Yandex Music A/B test.

  6. Why Post-Norm Transformers Collapse: Attention Amplification and Gradient Repair Failure

    cs.LG 2026-08 conditional novelty 6.0 of 10

    Token similarity in Post-Norm decoders is amplified by causal attention at initialization, and RMSNorm backward contraction prevents gradients from repairing the resulting collapse.

  7. Surface Keypoint Representation for Multi-Object and Articulated Human-Object Interaction Generation

    cs.CV 2026-08 conditional novelty 6.0 of 10

    Surface keypoint trajectories plus a spatio-temporal contact distance field let one generative model synthesize human motion for single, multiple, and articulated objects without specifying joint types.

  8. Raven: High-Recall Sequence Modeling with Sparse Memory Routing

    cs.LG 2026-07 conditional novelty 6.0 of 10

    Raven is a linear-time sequence model whose sparse, input-dependent routing writes tokens into dedicated memory slots, preserving long-context recall and extrapolating 16x beyond training length.

  9. GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation

    cs.CV 2026-03 conditional novelty 6.0 of 10

    A causal transformer with 3D RoPE generates vector-quantized 3D Gaussian latent grids autoregressively, enabling unconditional synthesis, completion, and open-ended outpainting of indoor scenes.

  10. Multi-view Pyramid Transformer: Look Coarser to See Broader

    cs.CV 2025-12 conditional novelty 6.0 of 10

    MVP uses a two-level hierarchy of attention windows and token resolutions to reconstruct large 3D scenes from up to 256 input views in a single feed-forward pass, beating Long-LRM and iLRM on DL3DV and several zero-sh...

  11. FLOWER: Democratizing Generalist Robot Policies with Efficient Vision-Language-Action Flow Policies

    cs.RO 2025-09 conditional novelty 6.0 of 10

    A compact 950-million-parameter robot policy trained in about 200 GPU-hours matches or beats multi-billion-parameter baselines on most manipulation benchmarks, including a new best score on CALVIN ABC.

  12. iLRM: An Iterative Large 3D Reconstruction Model

    cs.CV 2025-07 conditional novelty 6.0 of 10

    iLRM reconstructs 3D Gaussian scenes from multiple photos through iterative refinement of viewpoint tokens, achieving higher quality and speed than prior feed-forward models.

  13. NeoBabel: A Multilingual Open Tower for Visual Generation

    cs.CL 2025-07 conditional novelty 6.0 of 10

    A 2B multilingual text-to-image model trained on 124M translated pairs matches or beats larger English-only baselines on English while scoring higher on the authors' multilingual benchmark extensions.

  14. Towards Foundation Models for Experimental Readout Systems Combining Discrete and Continuous Data

    cs.LG 2025-05 conditional novelty 6.0 of 10

    A single transformer with separate pixel and time vocabularies generates realistic Cherenkov detector hits and supports particle identification and noise filtering after fine-tuning.

  15. Fast Text-to-Audio Generation with Adversarial Post-Training

    cs.SD 2025-05 conditional novelty 6.0 of 10

    ARC post-training speeds up text-to-audio generation to near-real-time speeds on GPUs and a few seconds on phones, without distillation or classifier-free guidance.

  16. RayZer: A Self-supervised Large View Synthesis Model

    cs.CV 2025-05 conditional novelty 6.0 of 10

    A self-supervised transformer model predicts camera poses and scene features from unposed images and renders novel views, reaching performance on par with pose-supervised baselines.

  17. Streaming DiLoCo with overlapping communication: Towards a Distributed Free Lunch

    cs.CL 2025-01 conditional novelty 6.0 of 10

    Streaming DiLoCo trains billion-parameter LLMs at data-parallel quality while cutting the inter-datacenter bandwidth by about two orders of magnitude via partial, overlapped, and 4-bit-quantized synchronization.

  18. DrivingWorld: Constructing World Model for Autonomous Driving via Video GPT

    cs.CV 2024-12 conditional novelty 6.0 of 10

    A 1B-parameter autoregressive world model generates over 40 seconds of controllable driving video with near-SOTA FVD, but key claims rest on inconsistent and cross-paper comparisons.

  19. NeuralPLexer3: Accurate Biomolecular Complex Structure Prediction with Flow Models

    cs.LG 2024-12 conditional novelty 6.0 of 10

    A flow-based model, NeuralPLexer3, predicts biomolecular complex structures with improved physical validity and speed over AlphaFold3 on the PoseBusters benchmark, and introduces NPBench and ConfBench for broader evaluation.

  20. SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training

    cs.CV 2024-12 conditional novelty 6.0 of 10

    SnapGen is a 379M-parameter UNet with cross-architecture distillation and a 1.38M-parameter decoder that generates 1024x1024 images on a phone in about 1.4 seconds, with GenEval 0.66 and ImageNet FID 2.06.

  21. BCAT: A Block Causal Transformer for PDE Foundation Models for Fluid Dynamics

    cs.LG 2025-01 conditional novelty 5.5 of 10

    BCAT, a block causal transformer for next-frame prediction, achieves state-of-the-art accuracy on 2D fluid dynamics PDE benchmarks, beating larger foundation models with fewer parameters.

  22. Next Token Prediction Towards Multimodal Intelligence: A Comprehensive Survey

    cs.CL 2024-12 conditional novelty 4.0 of 10

    A comprehensive survey that frames multimodal understanding and generation as next token prediction and proposes a five-part taxonomy.

Pith tools