HumanEval

Timeline

1. Oct 19, 2026 · Opinion / analysis · 1 source
   Qwen 3.8 Flash Next optimized for two CMP 170hx GPUs
   I've been working on optimizing Qwen 3.8 Flash Next on my local setup with two CMP 170HX cards and finally got to a place where I feel like the results are worth sharing.
2. Oct 18, 2026 · Opinion / analysis · 1 source
   Agens Volundr 32B v1: we fixed what you flagged in the Preview. 0/50 loops, session continuation, and an MIT engine that runs it with 262K context (measured to 239K) within a 16 GB budget
   WHAT v1 FIXES FROM THE PREVIEW
3. Oct 17, 2026 · Opinion / analysis · 1 source
   Xolotl (Strata based) - DeepSeek-V4.1-Flash (552B) on a single 4090 at ~27 tok/s, Uncensored 1017 quant coming in a couple of days
   I've been building an inference engine based off Strata called Xolotl that runs DeepSeek-V4.1-Flash, all 552B of it, on one gaming PC.
4. Oct 16, 2026 · Open-source release · 2 sources
   [Research] 2,788 tokens/s on a single RTX 5090 with a diffusion LLM — Introducing ORL (fully open source)
   We're excited to share ORL (On-Policy Residual Learning), our new research on making diffusion language models (dLLMs) significantly faster without sacrificing generation quality.
5. Oct 16, 2026 · Research paper · 1 source
   Next Token or Vector? Learning How Far to Decode
   We propose CADENCE (Context-Adaptive DEcoding), a generation method that learns both what to draft and how far to draft.
6. Oct 14, 2026 · Research paper · 1 source
   Decoupled Marginal Sharpening for Training-Free Inference-Time Scaling
   We study how to improve LLM reasoning at inference time without updating model weights or using external reward models.
7. Oct 13, 2026 · Research paper · 1 source
   Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference
   We present the first unified evaluation of weight-only quantization strategies for LLM inference across varying input/output lengths and GPUs, highlighting differences in energy efficiency in addition to accuracy degradation and runtime.
8. Oct 12, 2026 · Research paper · 1 source
   PIG: Parallel Inconsistency Guidance for Accelerating Diffusion Language Models
   Diffusion language models accelerate generation by decoding multiple masked positions per forward pass.
9. Oct 11, 2026 · Opinion / analysis · 1 source
   I tested different Qwen 3.8 27B quants
   There is a lot of discussion which quant to use.
10. Oct 11, 2026 · Model release · 1 source
    OrcaRouter Releases OrcaCyber Zero 1.5 Cybersecurity Model With 1M Context
    OrcaRouter has released OrcaCyber Zero 1.5, a model for authorized vulnerability research.
11. Oct 10, 2026 · Research paper · 1 source
    Beyond Adjacent Layers: Graph-Guided Layer Fusion for Compressing Large Language Models
    Existing depth-compression methods for large language models (LLMs) score layers individually or group only adjacent or contiguous layers, implicitly assuming that functional redundancy is local.
12. Oct 10, 2026 · Research paper · 1 source
    The Optimal Decoder is Tilted Temperature
    Larger sample budgets benefit from higher sampling temperatures in LLMs, but what temperature tuning optimizes and why it works remain poorly understood.
13. Oct 9, 2026 · Research paper · 1 source
    QTestBench: A Test Framework for Evaluating LLM-Generated Quantum Code
    We introduce QTestBench, a framework that evaluates LLM-generated quantum test suites against independent benchmark checkers, never treating the tests themselves as ground truth.
14. Oct 8, 2026 · Research paper · 1 source
    STAPLE: Static Sparse Upcycling of Pretrained Checkpoints via Embedding Modules
    In this work, we introduce STAPLE, a recipe built upon STEM for integrating static sparse memory into pretrained models.
15. Oct 8, 2026 · Research paper · 1 source
    Conditional Transfer from Controlled Pretraining Mixtures to Code
    Synthetic tasks are increasingly used both as probes of language-model capability and as pretraining data.
16. Oct 8, 2026 · Research paper · 1 source
    From Chain-of-Thought to Loops: Non-Autoregressive Latent Reasoning via Looped Transformers
    We introduce LLoCoT: a looped latent-reasoning framework that replaces left-to-right latent generation with iterative refinement of a compact latent workspace.
17. Oct 6, 2026 · Research paper · 1 source
    Few Bits, One Law: Toward W2A4KV2
    We introduce CanonQ, a unified quantization-aware training framework that addresses these challenges by separating source canonicalization from task-aware adaptation.
18. Oct 6, 2026 · Research paper · 1 source
    Adaptive Power Sampling for LLM Reasoning
    Sequence-level power sampling has recently emerged as a training-free approach to reasoning by sampling from a sharpened output distribution of a base large language model (LLM).
19. Oct 6, 2026 · Benchmark result · 1 source
    Qwen3.8 Flash Next Reasoning Modes: Off vs Low vs Medium vs Xhigh
    Like Qwen3.8 27B, Qwen3.8 Flash Next has three reasoning efforts: low, medium, and xhigh.

Often appears with