Memory Inception is a training-free method that injects latent KV banks at chosen layers to steer LLMs, achieving superior control-drift balance and up to 118x storage reduction on personality and structured-reasoning tasks.
and Zhang, Hao and Stoica, Ion , booktitle =
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
years
2026 7verdicts
UNVERDICTED 7representative citing papers
FoE restructures MoE blocks into per-KV-head clusters with sum-based synchronization, removing all-to-all communication in single-node settings and limiting it to intra-node in multi-node settings for up to 5.2x faster inference with comparable quality.
Feather uses reinforcement learning and a Chunked Hash Tree to balance batch size against prefix homogeneity in LLM inference, delivering 2-10x higher throughput than existing schedulers.
SynConfRoute routes code completions using syntax validation and token confidence, improving pass@1 by up to 31% on hard tasks and reducing accelerator usage by 58% versus always using the largest model.
HELM adaptively partitions HBM between EMB and KV caches via a three-layer PPO controller and EMB-KV-aware scheduling, reducing P99 latency by 24-38% while achieving 93.5-99.6% SLO satisfaction on production workloads.
PARSE accelerates LLM inference via parallel semantic prefix verification in a single forward pass, delivering 1.25x-4.3x speedups alone and up to 4.5x when combined with EAGLE-3.
AAFLOW is a unified distributed runtime that models agentic workflows as operators with a zero-copy data plane using Apache Arrow and Cylon, achieving up to 4.64x pipeline speedup through improved data flow and batching.
citing papers explorer
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception is a training-free method that injects latent KV banks at chosen layers to steer LLMs, achieving superior control-drift balance and up to 118x storage reduction on personality and structured-reasoning tasks.
-
Federation of Experts: Communication Efficient Distributed Inference for Large Language Models
FoE restructures MoE blocks into per-KV-head clusters with sum-based synchronization, removing all-to-all communication in single-node settings and limiting it to intra-node in multi-node settings for up to 5.2x faster inference with comparable quality.
-
Requests of a Feather Must Flock Together: Batch Size vs. Prefix Homogeneity in LLM Inference
Feather uses reinforcement learning and a Chunked Hash Tree to balance batch size against prefix homogeneity in LLM inference, delivering 2-10x higher throughput than existing schedulers.
-
SynConfRoute: Syntax-Aware Routing for Efficient Code Completion with Small CodeLLMs
SynConfRoute routes code completions using syntax validation and token confidence, improving pass@1 by up to 31% on hard tasks and reducing accelerator usage by 58% versus always using the largest model.
-
One Pool, Two Caches: Adaptive HBM Partitioning for Accelerating Generative Recommender Serving
HELM adaptively partitions HBM between EMB and KV caches via a three-layer PPO controller and EMB-KV-aware scheduling, reducing P99 latency by 24-38% while achieving 93.5-99.6% SLO satisfaction on production workloads.
-
Parallel Prefix Verification for Speculative Generation
PARSE accelerates LLM inference via parallel semantic prefix verification in a single forward pass, delivering 1.25x-4.3x speedups alone and up to 4.5x when combined with EAGLE-3.
-
AAFLOW: Scalable Patterns for Agentic AI Workflows
AAFLOW is a unified distributed runtime that models agentic workflows as operators with a zero-copy data plane using Apache Arrow and Cylon, achieving up to 4.64x pipeline speedup through improved data flow and batching.