# RWKV-7 "Goose" with Expressive Dynamic State Evolution

arXiv: https://arxiv.org/abs/2503.14456 · HTML: https://mlsanity.com/paper/2503.14456/

## Key facts

- Cited by 141 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Mar 2025.

## Built on by (most cited first, 30 of 141)

- [Kimi Linear: An Expressive, Efficient Attention Architecture](https://mlsanity.com/paper/2510.26692/) (2025)
- [Kimi K3: Open Frontier Intelligence](https://mlsanity.com/paper/2607.24653/) (2026)
- [A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyond](https://mlsanity.com/paper/2503.21614/) (2025)
- [Nested Learning: The Illusion of Deep Learning Architectures](https://mlsanity.com/paper/2512.24695/) (2025)
- [ATLAS: Learning to Optimally Memorize the Context at Test Time](https://mlsanity.com/paper/2505.23735/) (2025)
- [It's All Connected: A Journey Through Test-Time Memorization, Attentional Bias, Retention, and Online Optimization](https://mlsanity.com/paper/2504.13173/) (2025)
- [Glyph: Scaling Context Windows via Visual-Text Compression](https://mlsanity.com/paper/2510.17800/) (2025)
- [Jet-Nemotron: Efficient Language Model with Post Neural Architecture Search](https://mlsanity.com/paper/2508.15884/) (2025)
- [Log-Linear Attention](https://mlsanity.com/paper/2506.04761/) (2025)
- [PaTH Attention: Position Encoding via Accumulating Householder Transformations](https://mlsanity.com/paper/2505.16381/) (2025)
- [Understanding Transformer from the Perspective of Associative Memory](https://mlsanity.com/paper/2505.19488/) (2025)
- [Evolution Strategies at the Hyperscale](https://mlsanity.com/paper/2511.16652/) (2025)
- [RADLADS: Rapid Attention Distillation to Linear Attention Decoders at Scale](https://mlsanity.com/paper/2505.03005/) (2025)
- [Efficient Attention Mechanisms for Large Language Models: A Survey](https://mlsanity.com/paper/2507.19595/) (2025)
- [Lattice: Learning to Efficiently Compress the Memory](https://mlsanity.com/paper/2504.05646/) (2025)
- [Structured Linear CDEs: Maximally Expressive and Parallel-in-Time Sequence Models](https://mlsanity.com/paper/2505.17761/) (2025)
- [Hybrid Linear Attention Done Right: Efficient Distillation and Effective Architectures for Extremely Long Contexts](https://mlsanity.com/paper/2601.22156/) (2026)
- [TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecasting](https://mlsanity.com/paper/2510.25502/) (2025)
- [Structured Sparse Transition Matrices to Enable State Tracking in State-Space Models](https://mlsanity.com/paper/2509.22284/) (2025)
- [Comba: Improving Bilinear RNNs with Closed-loop Control](https://mlsanity.com/paper/2506.02475/) (2025)
- [Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models](https://mlsanity.com/paper/2511.18890/) (2025)
- [Hybrid Architectures for Language Models: Systematic Analysis and Design Insights](https://mlsanity.com/paper/2510.04800/) (2025)
- [Scaling Linear Attention with Sparse State Expansion](https://mlsanity.com/paper/2507.16577/) (2025)
- [Fast-weight Product Key Memory](https://mlsanity.com/paper/2601.00671/) (2026)
- [Test-Time Training with KV Binding Is Secretly Linear Attention](https://mlsanity.com/paper/2602.21204/) (2026)
- [Gated KalmaNet: A Fading Memory Layer Through Test-Time Ridge Regression](https://mlsanity.com/paper/2511.21016/) (2025)
- [InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models](https://mlsanity.com/paper/2512.08829/) (2025)
- [AuroraLong: Bringing RNNs Back to Efficient Open-Ended Video Understanding](https://mlsanity.com/paper/2507.02591/) (2025)
- [Selective Rotary Position Embedding](https://mlsanity.com/paper/2511.17388/) (2025)
- [RAD: Redundancy-Aware Distillation for Hybrid Models via Self-Speculative Decoding](https://mlsanity.com/paper/2505.22135/) (2025)

## Abstract (arXiv)

We present RWKV-7 "Goose", a new sequence modeling architecture with constant memory usage and constant inference time per token. Despite being trained on dramatically fewer tokens than other top models, our 2.9 billion parameter language model achieves a new 3B SoTA on multilingual tasks and matches the current 3B SoTA on English language downstream performance. RWKV-7 introduces a newly generalized formulation of the delta rule with vector-valued gating and in-context learning rates, as well as a relaxed value replacement rule. We show that RWKV-7 can perform state tracking and recognize all regular languages, while retaining parallelizability of training. This exceeds the capabilities of Transformers under standard complexity conjectures, which are limited to mathsfTC0. To demonstrate RWKV-7's language modeling capability, we also present an extended open source 3.1 trillion token multilingual corpus, and train four RWKV-7 models ranging from 0.19 billion to 2.9 billion parameters on this dataset. To foster openness, reproduction, and adoption, we release our models and dataset component listing at https://huggingface.co/RWKV, and our training and inference code at https://github.com/RWKV/RWKV-LM all under the Apache 2.0 License.