arXiv:2609.00715v1 Announce Type: cross Abstract: Feedback-based coding schemes have demonstrated substantial performance gains over today's open-loop coding schemes. Unfortunately, these gains are u...
arXiv:2608. 04405v1 Announce Type: cross Abstract: Long-context large language models (LLMs) are increasingly deployed in real-world applications, yet self-attention remains a major efficiency bottleneck -- especially during decoding -- due to the necessity of repeatedly processing ever-growing key-value (KV) caches.
Long-context large language models (LLMs) are increasingly deployed in real-world applications, yet self-attention remains a major efficiency bottleneck -- especially during decoding -- due to the necessity of repeatedly processing ever-growing key-value (KV) caches. Existing sparse attention reduce computation by attending to fewer KV pairs, but often suffer from substantial accuracy degradation, require additional training, or rely on expensive hashing.
MambaCSP is a hybrid-attention state space model that replaces transformer-based backbones with a linear-time Mamba architecture for channel state prediction. By adding lightweight patch‑mixer attention layers, it captures long‑range dependencies while maintaining hardware efficiency. Experiments on MISO‑OFDM show 9‑12% higher accuracy, 3× faster throughput, 2.6× lower VRAM usage, and 2.9× faster inference compared to LLM‑based methods.
arXiv:2608. 02032v1 Announce Type: new Abstract: Modern language models are built primarily from Transformers, recurrent models, and their hybrid architectures.
The paper introduces Memory-Augmented Source Coding (MASC), a scheme that embeds contextual patterns into a source model to improve robustness in low‑SNR communications. MASC uses a shared Parameterized Contextual Memory (PCM) for multi‑order n‑gram patterns and a Mixture‑of‑Memory‑Experts Router (MMER) to selectively activate memory experts based on hidden states, thereby refining probability estimates and shortening code length. Experiments on Rayleigh fading and AWGN channels show that MASC reduces decoding sensitivity to residual channel errors compared to traditional SSCC with autoregressive decoding and LLM‑based Arithmetic Coding.
arXiv:2607. 07953v1 Announce Type: cross Abstract: Self-attention lets each token retrieve information from the full context, but its quadratic cost in sequence length limits training and inference at long context.
arXiv:2608. 14191v1 Announce Type: new Abstract: The key-value (KV) cache stores information from past tokens and is a major memory bottleneck in long-context inference.
Modern language models are built primarily from Transformers, recurrent models, and their hybrid architectures. Transformers rely on token-level attention memories, while recurrent models such as state space models (SSMs) and linear attention maintain compact recurrent states.
arXiv:2602. 07223v2 Announce Type: replace Abstract: Long-context large language model (LLM) inference has become the norm for today's AI applications.
arXiv:2608. 02691v1 Announce Type: cross Abstract: The key-value (KV) cache has become a major memory and bandwidth bottleneck in long-context large language model inference, making ultra-low-bit quantization increasingly important.
arXiv:2606. 31519v1 Announce Type: new Abstract: Long-context Large Language Model inference is severely bottlenecked by the massive Key-Value (KV) cache, yet existing sparse attention methods often suffer from static fixed-budget (Top-k) retrieval or rely on proxy scores that are computationally expensive and biased.