Loom implements a 22-opcode C-compatible computer inside an 8-layer transformer with analytically derived, program-independent weights, executing instructions iteratively on a fixed-size state tensor at constant per-step cost.
Autoregressive large language models are computationally universal
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
Introduces TSPD with a trajectory-feature controller and training-free CE to reduce denoising steps in dLLMs while aiming to preserve quality.
Existing proofs of autoregressive Transformer Turing-completeness apply to scaling families of models rather than fixed systems with context management, so they do not establish Turing-completeness for real-world LLMs.
High-dimensional geometry imposes concurrency limits on semantic directions in LLM embeddings via residual interference, with N < exp(c d_eff ε²) for coexistence and σ_int ~ √(k/d_eff) for readout noise.
citing papers explorer
-
Loom: A Scalable Analytical Neural Computer Architecture
Loom implements a 22-opcode C-compatible computer inside an 8-layer transformer with analytically derived, program-independent weights, executing instructions iteratively on a fixed-size state tensor at constant per-step cost.
-
Efficient Diffusion LLMs via Temporal-Spatial Parallel Decoding and Confidence Extrapolation
Introduces TSPD with a trajectory-feature controller and training-free CE to reduce denoising steps in dLLMs while aiming to preserve quality.
-
Position: The Turing-Completeness of Autoregressive Transformers Relies Heavily on Context Management
Existing proofs of autoregressive Transformer Turing-completeness apply to scaling families of models rather than fixed systems with context management, so they do not establish Turing-completeness for real-world LLMs.