rwkv

The author introduces Helix-AGI, an open-source agent harness that replaces JSON tool-calling with a dual-RWKV bicameral architecture pairing a conversational Speaking Head with a non-language Sensory-Actualizer Head, coupled in latent space and driven by a Habitus graph of learned habitual behaviors with real-time visualization.

论文提出 HART，一种混合注意力与 RWKV 的时间序列基础模型，通过交替注意力和 RWKV 模块解决非自回归多步预测中 RWKV-7 的长程退化问题（Placeholder Idling Decay）。仅用 36M 参数（约为 Toto-2.0-1B 的 1/28）在 GIFT-Eval 上达到可媲美的长程 MASE（0.752 vs 0.754）。

NinaXander proposes composing frozen language models from different architecture families (Transformer and RWKV) via a single shared-latent adapter, achieving an 84.4% KV-cache reduction without accuracy loss to RWKV alone, though composed models still underperform their strongest parent and degrade out-of-domain.

The article compares the routes of two AI model architecture camps, Mamba and RWKV: Mamba targets cloud-side enterprise AI platforms, while RWKV targets on-device and edge devices, each adapting to different hardware and inference scenarios.

RWKV is a novel language model architecture that combines the efficiency of RNNs with the parallelizability of Transformers, achieving strong LLM performance with linear time complexity and constant memory footprint.

BlinkDL announces RWKV-7 G1g, a pure RNN LLM that claims to be the best in its class and competitive with general LLMs, with high-speed inference on a single RTX 5090.

Ali Hatamizadeh announces Gated DeltaNet-2, a new linear attention model that outperforms KDA and Mamba-3 at 1.3B scale; @BlinkDL_AI notes its recurrence is nearly identical to RWKV-7's DPLR.