# LLM-42: Enabling Determinism in LLM Inference with Verified Speculation

arXiv: https://arxiv.org/abs/2601.17768 · HTML: https://mlsanity.com/paper/2601.17768/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 62 references that are in the map.
- First posted on arXiv 25 Jan 2026.

## Builds on (most cited first, 30 of 33)

- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning](https://mlsanity.com/paper/2307.08691/) (2023)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [Accelerating Large Language Model Decoding with Speculative Sampling](https://mlsanity.com/paper/2302.01318/) (2023)
- [Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads](https://mlsanity.com/paper/2401.10774/) (2024)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU](https://mlsanity.com/paper/2303.06865/) (2023)
- [Splitwise: Efficient generative LLM inference using phase splitting](https://mlsanity.com/paper/2311.18677/) (2023)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision](https://mlsanity.com/paper/2407.08608/) (2024)
- [A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents](https://mlsanity.com/paper/1804.05685/) (2018)
- [SpecInfer: Accelerating Generative Large Language Model Serving with Tree-based Speculative Inference and Verification](https://mlsanity.com/paper/2305.09781/) (2023)
- [LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding](https://mlsanity.com/paper/2404.16710/) (2024)
- [Break the Sequential Dependency of LLM Inference Using Lookahead Decoding](https://mlsanity.com/paper/2402.02057/) (2024)
- [SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefills](https://mlsanity.com/paper/2308.16369/) (2023)
- [Inference without Interference: Disaggregate LLM Inference for Mixed Downstream Workloads](https://mlsanity.com/paper/2401.11181/) (2024)
- [The Good, The Bad, and The Greedy: Evaluation of LLMs Should Not Ignore Non-Determinism](https://mlsanity.com/paper/2407.10457/) (2024)
- [DynamoLLM: Designing LLM Inference Clusters for Performance and Energy Efficiency](https://mlsanity.com/paper/2408.00741/) (2024)
- [vAttention: Dynamic Memory Management for Serving LLMs without PagedAttention](https://mlsanity.com/paper/2405.04437/) (2024)
- [NanoFlow: Towards Optimal Large Language Model Serving Throughput](https://mlsanity.com/paper/2408.12757/) (2024)
- [Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts](https://mlsanity.com/paper/2502.19811/) (2025)
- [POD-Attention: Unlocking Full Prefill-Decode Overlap for Faster LLM Inference](https://mlsanity.com/paper/2410.18038/) (2024)
- [Dynamic Speculation Lookahead Accelerates Speculative Decoding of Large Language Models](https://mlsanity.com/paper/2405.04304/) (2024)
- [TileLink: Generating Efficient Compute-Communication Overlapping Kernels using Tile-Centric Primitives](https://mlsanity.com/paper/2503.20313/) (2025)
- [TokenWeave: Efficient Compute-Communication Overlap for Distributed LLM Inference](https://mlsanity.com/paper/2505.11329/) (2025)
- [Deterministic Inference across Tensor Parallel Sizes That Eliminates Training-Inference Mismatch](https://mlsanity.com/paper/2511.17826/) (2025)
- [Efficient and Adaptable Overlapping for Computation and Communication via Signaling and Reordering](https://mlsanity.com/paper/2504.19519/) (2025)
- [A Comprehensive Analysis of Large Language Model Outputs: Similarity, Diversity, and Bias](https://mlsanity.com/paper/2505.09056/) (2025)
- [Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding](https://mlsanity.com/paper/2512.01278/) (2025)

## Built on by (most cited first, 16 of 16)

- [Watt Counts: Energy-Aware Benchmark for Sustainable LLM Inference on Heterogeneous GPU Architectures](https://mlsanity.com/paper/2604.09048/) (2026)
- [VibeServe: Can AI Agents Build Bespoke LLM Serving Systems?](https://mlsanity.com/paper/2605.06068/) (2026)
- [Decoding as Optimisation on the Probability Simplex: From Top-K to Top-P (Nucleus) to Best-of-K Samplers](https://mlsanity.com/paper/2602.18292/) (2026)
- [Demystifying Numerical Instability in LLM Inference: Achieving Reproducible Inference for Mission-Critical Tasks with HEAL](https://mlsanity.com/paper/2606.21023/) (2026)
- [The Integer Alibi: Localizing Cross-Kernel Divergence in INT8-Quantized LLM Inference](https://mlsanity.com/paper/2608.13756/) (2026)
- [Accelerating the Mitigation of LLM Inference Nondeterminism Across GPU Architectures](https://mlsanity.com/paper/2609.25624/) (2026)
- [From Accuracy to Auditability: A Survey of Determinism in Financial AI Systems](https://mlsanity.com/paper/2605.23955/) (2026)
- [A Paired Testing Protocol for Batch-Conditioned Refusal Robustness in LLM Serving](https://mlsanity.com/paper/2605.27763/) (2026)
- [Reproducibility is the New Copyleft: Defining AGI-oriented Reproducible Builds](https://mlsanity.com/paper/2606.03019/) (2026)
- [Speculative Decoding at Temperature Zero: A Scoped Safety-Invariance Screen with a 48,072-Sample Expansion](https://mlsanity.com/paper/2606.25097/) (2026)
- [Same Request, Different Answer: Quantization Amplifies Cache-Induced Divergence in LLM Serving](https://mlsanity.com/paper/2609.04748/) (2026)
- [Validating Hybrid-State Cache Recovery for GLM-5.3-Flash with vLLM and LMCache](https://mlsanity.com/paper/2609.15030/) (2026)
- [Greedy Decoding Is Not Precision-Invariant: Cross-Precision Output Divergence in LLM Inference](https://mlsanity.com/paper/2609.26621/) (2026)
- [When Do Models Admit They Are Wrong? Failure Disclosure Is Unstable Under Reinforcement Learning](https://mlsanity.com/paper/2609.33220/) (2026)
- [Measuring and Reducing Cross-Vendor Mismatch in Language Models](https://mlsanity.com/paper/2610.05458/) (2026)
- [Same Text, Different Prediction: Serving-Context Nondeterminism in Text Classifiers](https://mlsanity.com/paper/2610.09111/) (2026)

## Abstract (arXiv)

In LLM inference, the same prompt may yield different outputs across different runs. At the system level, this non-determinism arises from floating-point non-associativity combined with dynamic batching and GPU kernels whose reduction orders vary with batch size. A straightforward way to eliminate non-determinism is to disable dynamic batching during inference, but doing so severely degrades throughput. Another approach is to make kernels batch-invariant; however, this tightly couples determinism to kernel design, requiring new implementations. This coupling also imposes fixed runtime overheads, regardless of how much of the workload actually requires determinism. Inspired by ideas from speculative decoding, we present LLM-42, a scheduling-based approach to enable determinism in LLM inference. Our key observation is that if a sequence is in a consistent state, the next emitted token is likely to be consistent even with dynamic batching. Moreover, most GPU kernels use shape-consistent reductions. Leveraging these insights, LLM-42 decodes tokens using a non-deterministic fast path and enforces determinism via a lightweight verify-rollback loop. The verifier replays candidate tokens under a fixed-shape reduction schedule, commits those that are guaranteed to be consistent across runs, and rolls back those violating determinism. LLM-42 mostly re-uses existing kernels unchanged and incurs overhead only in proportion to the traffic that requires determinism.