SimLens is a training‑free decoder that improves early‑layer predictions in large language models by keeping only the start token and a candidate answer token and performing a lightweight continuation through the remaining layers. It outperforms direct linear readouts, yielding higher accuracy on tasks such as ARC, BoolQ, and HeadQA with LLaMA‑7B and Vicuna‑7B. The method is extended to Linear SimLens for confidence estimation and combined into SimExit, a hybrid early‑exit mechanism that achieves significant speedups while maintaining accuracy.
arXiv:2609.36590v1 Announce Type: cross Abstract: Self-speculative decoding accelerates large language model (LLM) inference by drafting tokens from the target model itself, but faces a sharp tradeof...
The paper introduces Variance‑Calibrated Modulation (VCM), a training‑free pre‑decoding technique that reshapes language model probability distributions before truncation. VCM uses two dynamic mechanisms: a Contextual Searchlight via PMI to suppress stopwords and highlight context‑relevant tokens, and an Adaptive Self‑Debiasing that applies scale‑invariant penalization based on real‑time logit standard deviation. Experiments on open‑ended generation, factual QA, and mathematical reasoning show that VCM consistently reduces the likelihood trap, improving diversity, coherence, and reasoning accuracy with minimal computational cost.
arXiv:2407. 21082v3 Announce Type: replace-cross Abstract: This paper presents a modular approach to accelerate inference in large language models (LLMs) by adding early exit heads at intermediate transformer layers.
Manacá-1B is a 1.72‑billion‑parameter, open decoder‑only language model trained from scratch for Brazilian Portuguese, released with a fully containerized, reproducible training pipeline and complete logs. The authors evaluate it against nine open baselines on four Portuguese benchmarks, reporting standard errors and paired significance tests, and find that Manacá-1B outperforms smaller models on LAMBADA‑PT while remaining competitive on commonsense completion. They also uncover a tokenizer‑related evaluation pitfall that can drastically lower accuracy and provide a simple fix, releasing all code, logs, and corrected tokenizer for full reproducibility.
arXiv:2606. 16847v1 Announce Type: cross Abstract: Diffusion Large Language Models (dLLMs) offer a promising avenue for parallel generation but face a trade-off between decoding speed and quality.
arXiv:2606. 02544v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) LLMs, offering faster inference through parallel or blockwise decoding.
arXiv:2607. 01792v1 Announce Type: cross Abstract: While decoder-only LLMs excel at a vast array of natural language tasks, it suffers from an asymmetric information flow induced by causal attention: later tokens are richer in contextual grounding than earlier ones.
The paper introduces Reliable Parallel Decoding (RPD) for masked diffusion language models, addressing the unreliability of committing multiple predictions from a single forward pass. Diagnostics reveal that confidence alone is insufficient, as confident end‑sequence predictions can preempt necessary upstream computations, and downstream predictions degrade with upstream uncertainty. RPD selects candidates based on layer‑wise stability and final confidence, committing them under an entropy budget while deferring uncertain predictions, achieving superior throughput and competitive accuracy on LLaDA and Dream benchmarks.
arXiv:2604. 18995v2 Announce Type: replace-cross Abstract: Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to autoregressive generation by enabling parallel token prediction.
The paper introduces DCOCO (DFlash Contextual Corrector), a lightweight mechanism that restores intra-block dependencies in speculative decoding by evaluating multiple candidate tokens for contextual compatibility across adjacent draft positions. DCOCO maintains parallel drafting while enabling contextually informed token selection at low additional cost. Across multiple verifier backbones and benchmarks, DCOCO achieves the highest average throughput, delivering speedups of up to 6.6× over vanilla decoding and outperforming competing methods by 6–23% under concurrent serving loads.
CoDR (Confidence Drift Remasking) is a training‑free, sampler‑agnostic refinement pass for masked diffusion language models that detects and corrects early token commitments that become unsupported as context grows denser. It estimates confidence drift for all committed positions using only a few forward passes and remasks only the tokens whose confidence has dropped, thereby improving accuracy across multiple backbones, tasks, and samplers with modest computational overhead.
Ripple-Pivot Search (RPS) is a training‑free decoding method for Diffusion Large Language Models that identifies mid‑entropy pivot positions to reduce uncertainty across remaining masked tokens. By proactively committing these pivots and evaluating token assignments via lookahead, RPS enables more tokens to be unmasked in parallel, speeding up decoding. Experiments on three dLLMs and four reasoning/code‑generation benchmarks show 4–10× wall‑clock speedup over standard decoding, up to 18× with KV caching, while maintaining or improving generation quality.