循環語言模型的延遲進入策略
Late Entry in Looped Language Models
研究提出循環語言模型的延遲進入策略,讓新 token 跳過初始 passes,直接進入最後 K 個 passes,利用深層上下文 cache 加快推理。在 Ouro-2.6B(T = 4)上,Wake 於 tau = 0 達到 5.2x greedy-exact 加速;在 Huginn-3.5B 上,T = 64 時達到 8.4x。
Published on Nov 18
·
on Nov 19
Authors:
,
,
Abstract
Looped language models deepen reasoning at a compact parameter count by iterating a shared layer stack T times per token, turning recurrence depth directly into sequential decoding latency. Conventional early-exit decoding halts after initial passes without cache completion, truncating cached key-value states of past tokens to shallow depth. We study the opposite strategy, late entry, where new tokens skip initial passes and enter directly into the final K < T passes over a deep context cache. Once lightweight fine-tuning resolves hidden-state norm mismatch at the entry boundary, deep cached context substitutes for skipped token-local passes, allowing late entry to outperform matched early exit across model families and pass budgets. To turn this asymmetry into wall-clock speedups without stalling to build deep caches, candidate blocks verify through only the final K passes on the critical path while previously committed tokens complete passes 1 -> T alongside subsequent steps at zero extra serial calls. Projecting these finished T-pass wake states inside the existing output GEMM (Wake verification, governed by logit-margin tolerance tau in [0, inf]) unifies greedy-exact T-pass decoding (tau = 0, matching K = T block decoding within bf16 reduction order), margin-gated verification (tau > 0), and non-rollback lossy late entry (tau = inf) inside a single static CUDA graph. On Ouro-2.6B (T = 4), K = 2 Wake reaches 5.2x greedy-exact speedup at tau = 0 (+26% over K = 4 block decoding on GSM8K) and 6.8x at tau = inf (+16.9 points over trained early exit at matched speed). Scaling test-time recurrence across T in {8, 16, 32, 64} on Huginn-3.5B cuts K = T/4 rollbacks 21.5x with depth, driving greedy-exact (tau = 0) speedups to 6.1x (T = 32) and 8.4x (T = 64, +153% over K = 64 block) and tau = inf speedups to 10.6 to 12.5x.
View arXiv page View PDF Project page GitHub Add to collection
Get this paper in your agent:
hf papers read 2611.24576
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2611.24576 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2611.24576 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2611.24576 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co