AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：For most languages of the world, language model pre-training operates in a data-constrained regime where models must repeat their training data many times, degrading generalization. Two remedies exist: aggressive hyperparameter tuning such as high weight decay, and mixing in data from a high-resource auxiliary language to directly aid the low-resource target. While hyperparameter tuning regularizes the model by shrinking weights to restrict network capacity, auxiliary data mixing uses a tunable mixing ratio to expand the training distribution and diversify the training signal with new…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Learn how to deploy decision models with llama.cpp on Amazon SageMaker AI. This technical how-to serves OpenAI-compatible chat completions with the AWS Deep Learning Container for llama.cpp, then extends the runtime with SemIf to return probabilities over declared options on CPU or GPU.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.12642v1 Announce Type: new Abstract: Motion retargeting is a critical preprocessing step for transferring human motion data to humanoid robots. While existing methods are effective for single-character motions, interactions involving multiple characters and objects introduce additional contacts and spatial relationships that must be preserved. Retargeting is commonly formulated as a weighted multi-objective optimization problem, where conflicting objectives can require careful tuning of relative weights and can still compromise important motion characteristics. We introduce PriMoRetarget, a prioritized motion-retargeting framework that replaces relative objective weighting with an explicit priority hierarchy. Lower-priority updates are constrained to…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.12516v1 Announce Type: new Abstract: Interleaved multimodal in-context learning (ICL) lets frozen large vision-language models (LVLMs) acquire new tasks from a handful of image-text-answer demonstrations, yet it remains fragile: demonstrations bind to the wrong images, redundant exemplars crowd out informative ones, and noisy answers are copied verbatim into predictions. We present AVA, a training-free framework that organizes the demonstration pipeline of multimodal ICL into three composable stages. A Thread-Resolved Decomposer (TRD) segments the interleaved context into per-sample threads and re-stitches them with explicit boundaries, restoring image-text binding. An Influence-scored Demonstration Selector (IDS) estimates each candidate's marginal…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.12617v1 Announce Type: new Abstract: Geopolitical and strategic intelligence reports constitute a rich source of expert knowledge and have increasing value in the current political climate, but their heterogeneity and analytical complexity pose challenges for retrieval-augmented question answering. We introduce a benchmark for grounded question answering over 415 open-source geopolitical and strategic intelligence reports, comprising 5,525 document-grounded question-answer pairs spanning factual, inferential, comparative, and risk-oriented questions. We propose a document-aware generation pipeline that profiles reports before generating questions and validate dataset quality through expert annotation and an LLM-as-judge protocol. Using the benchmark,…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：TII has released Falcon-ASR, a 1.6B speech recognition model for Arabic with an Emirati focus. It reports 20.92% average WER on 6 Arabic test sets, but weights and an API are not yet available. The post TII Releases Falcon-ASR: 1.6B Arabic Speech Model Hits 20.92% WER appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Reflection and Mistral expand the model race, TypeSafe and Arena raise the stakes, and OpenAI puts mathematical discovery under scrutiny.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Alibaba’s Qwen team has released Qwen-Image-2.1-Turbo, an accelerated checkpoint of its open-weight Qwen-Image-2.1 model. It generates and edits images in 8 denoising steps instead of the base model’s 40-step default. For developers, that means 5x fewer denoising steps on the same 7B architecture, plus a hosted API option. TL;DR What is Qwen-Image-2.1-Turbo? Qwen-Image-2.1-Turbo is an […] The post Alibaba Qwen Releases Qwen-Image-2.1-Turbo, an 8-Step 7B Image Model appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：EmbeddingGemma 2 launched on October 6, 2026 under Apache 2.0. It is a sub-1B model built on Gemma 4 that maps text, code, images, video and audio into one 768-dimensional space. This article covers the architecture, the benchmarks, and runnable scripts to provide measured results. Specifications Specification EmbeddingGemma 2 Base model Gemma 4 License Apache […] The post EmbeddingGemma 2: Text, Code, Images, Video and Audio in One Vector Space appeared first on Analytics Vidhya.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Llama 3 is a capable generalist, but that's exactly the problem when you need an agent that reliably outputs structured JSON for a specific API.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：OpenAI publishes hundreds of math proofs from unreleased frontier model, Mistral and Reflection AI launch open-weight models to rival China, and more!
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10859v1 Announce Type: new Abstract: Text-to-image diffusion models are increasingly distilled into few-step variants and being deployed to enable fast inference. However, their ability to generate harmful or undesired content poses significant safety risks. Data-driven unlearning methods suppress targeted generations by fine-tuning model weights using specialized unlearning objectives. Crucially, these objectives implicitly rely on multi-step denoising dynamics, an assumption that breaks down for few-step distilled (FSD) models, resulting in ineffective forgetting. Furthermore, performing unlearning on the non-distilled base model and subsequently re-distilling it to obtain an unlearned FSD model incurs substantial computational and time overhead, m…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10845v1 Announce Type: new Abstract: A large language model can only use the text that fits in its context window, and it recomputes its internal key-value (KV) state for a prompt every time the prompt is sent. We test a memory layer, the public package galahad-kv, that saves the KV state of each block of about 16,000 tokens to encrypted local NVMe disk and loads it back later, byte-exact, without recomputing it. We ran it on 50,000,000 tokens of real public text, served through vLLM on one NVIDIA H100, with Gemma 4 12B and Gemma 4 31B. Every block we probed was loaded back from the encrypted store with no recompute (100 of 100, at depths from 0 to 50M tokens) on both models. Loading a block was 2.8x to 4.3x faster than recomputing it and used 8.8x t…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10827v1 Announce Type: new Abstract: Corrective feedback is among the best-evidenced drivers of second-language acquisition, yet corrections delivered during lessons rarely accumulate into an actionable view of grammar mastery. Prompted frontier models can provide such a view from learner--tutor lesson transcripts, but they are costly at scale. We close this gap by fine-tuning Qwen3.5 small language models (SLMs) on filtered and rebalanced teacher-generated supervision, then deploying an efficient 0.8B model in an end-to-end grammar mastery tracker for all English learners on our platform. Internalizing the annotation contract into adapter weights enables pairing the 0.8B model with a compact matched prompt rather than verbose instructions. On two hu…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10592v1 Announce Type: new Abstract: Historical Polish is well documented as a language but annotated in machine-readable form only to about a million words for the period this paper covers; the rest sits behind optical character recognition of variable quality. We present Wieszcz-XIX, a corpus of 6.75 billion tokens (about 3.1 billion words) in 294,369 documents, most of them periodical issues, of Polish published from 1800 to 1918, assembled from Wolne Lektury and the Internet Archive by a pipeline that filters, deduplicates, audits for post-1918 leakage and splits at the document level. It is over three orders of magnitude larger than the annotated corpus of the same period, and we quantify its defects: recognition corruption against a false-posit…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10550v1 Announce Type: new Abstract: Personalizing text-to-image diffusion models from a few reference images requires preserving subject identity while following prompts that describe new contexts. Full-model fine-tuning is parameter-intensive, whereas low-rank adaptation (LoRA) reduces the number of trainable parameters but leaves open how adaptation capacity should be distributed across layers. We introduce Diffu-LoRA, a parameter-efficient method that learns this allocation through gated low-rank adaptation. Diffu-LoRA inserts trainable low-rank components into the linear layers of Transformer blocks and assigns a learnable gate to each component. Bilevel optimization updates the adaptation weights and gate parameters on separate data splits, whi…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10594v1 Announce Type: new Abstract: Activation probes that monitor deployed language models are trained on synthetic conversations, and how many a probe needs is open. We trace learning curves over 10-590 synthetic samples for three monitoring concepts, high-stakes situations, replies harmful to a person, and replies that do not follow the user's instruction, on fourteen held-out evaluation distributions and four probe models, varying the generator LLM and the prompt's detail. The need is set by what is monitored: probes for high-stakes and harmful are within a few hundredths of their plateau from 80 samples on Gemma-3-27B-IT, instruction probes need several times as many, and the ordering holds on three smaller probe models and on real samples (fro…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10552v1 Announce Type: new Abstract: Comparing parameter-efficient fine-tuning recipes under a single, shared learning rate is a common but flawed practice: when the arms being compared have very different trainable-parameter counts, a shared rate can simultaneously depress the larger arms' means and inflate their variance, manufacturing a large, seemingly multi-seed-significant advantage for the smallest arm that is not a real effect. We document this confound in a concrete setting: post-hoc SVD-based KV-cache compression, where an already-pretrained model is converted to a low-rank (multi-head-latent-attention-style) cache by factorizing its key/value weights into a down-projection ("encoder") and an up-projection ("decoder"), after which a short f…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.10629v1 Announce Type: new Abstract: Self-improving LLM agents can adapt a credit pipeline to a changed rule, but an agent that rewrites itself destroys the artefact a supervisor reviews: a named change, a recorded test, an approval. We argue that self-evolution is reviewable only if it is confined to the runtime harness (instruction text, tool-call logic and primitive composition) while model weights stay fixed, so that every adaptation is a diff with a cause and a test attached. We give a dual-loop engine built on that bound, with one admission gate that writes a hash-chained record before deployment, and we measure the gate in simulation, with a simulated agent and a seeded-search proposer rather than language models. Across three families of supe…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：New Western open-weight models could give enterprises an alternative to Chinese and proprietary AI, but greater control also brings new responsibilities.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Google has released an experimental note-taking app that can transcribe meetings and audio files entirely offline, as reported earlier by TechCrunch. The app, called Google AI Edge Foresight, is free to use and runs on macOS using the company's on-device EmbeddingGemma 2 model. Similar to AI note-taking apps like Granola and Wispr Flow, Foresight summarizes your meetings while providing a space where you can take your own notes. Google says you can jot down shorthand bullet points during your meeting, and the app will transform them into "polished notes" based on the information in the transcript. Along with transcribing your meetings, For … Read the full story at The Verge.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.08858v1 Announce Type: new Abstract: Low-rank compression reduces the cost of pretrained language models by replacing linear transformations with low-rank factorizations. However, conventional methods use a fixed rank allocation during inference, assigning the same amount of compute regardless of the input token. We introduce Low-Rank Conditional Computation (LRCC), which adds token-dependent computation to pretrained models by training one lightweight router per Transformer block to select among a small set of nested low-rank paths. During training, the low-rank factors remain frozen, and only the routers are optimized. We evaluate LRCC on Llama and Qwen models for language modeling and zero-shot downstream tasks. Within the same average active-para…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.08794v1 Announce Type: new Abstract: Large language models rely on subword tokenizers whose quality varies across languages, yet no standardized multi-metric framework exists for broad comparative evaluation. We introduce Tokka-Bench, an open-source framework that evaluates tokenizers on five complementary metrics -- bytes per token, unique token coverage, subword fertility, word-split rate, and vocabulary composition -- across 100 natural languages (30+ scripts) and 20 programming languages, using language-aware segmentation adapted to each writing system. Comparing seven BPE tokenizers (GPT-2, GPT-4, gpt-oss, Llama 3.1, Gemma 3, Qwen3, and Kimi K2) within individual languages, we find that vocabulary allocation strategy matters more than raw vocabu…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Liquid AI has released Open d1, two open-weight multimodal models in its d1 decision model family. d1-3B reads text and images. d1-omni-600M reads text with an image, or text with audio. Neither model writes text. Each returns calibrated, typed answers in one forward pass with zero output tokens. The target is real-time decisions on the […] The post Liquid AI Releases Open-Weight d1-3B and d1-omni-600M: Multimodal Decision Models With Zero Output Tokens appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.07032v1 Announce Type: new Abstract: Large-scale pretrained transformer models have achieved state-of-the-art performance across diverse machine translation tasks, including multilingual settings. Knowledge distillation has emerged as a sustainable approach for model compression, transferring knowledge from large teacher models to smaller, more efficient student models. Similarly, quantization, which reduces the numerical precision of model weights and activations (e.g., from 32-bit to 8-bit representations) is widely used to accelerate inference, enabling models to run several times faster during deployment. However, both techniques face limitations when applied to specialized domain data, particularly under low-resource conditions. In knowledge dis…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.07019v1 Announce Type: new Abstract: Fiorillo v0.5 is an open model that answers typed questions with a probability for each answer. Its main specialist reads a randomized trial's article, cut to 6,144 tokens, and answers whether an intervention significantly increased, significantly decreased or did not significantly change an outcome against a comparator (Evidence Inference 2.0, EI). It is Qwen3-4B-Base with low-rank adapters and a decision head, fine-tuned for EI only on the 1,431 of 2,657 training articles whose own license allows reuse. Four criteria registered on the Open Science Framework before this version's test predictions decided its release, the second bar judged on EI's test split, whose labels are public. On that split (1,218 prompts i…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.06903v1 Announce Type: new Abstract: Activation steering manipulates large language model behavior by intervening on internal activations, but the mechanistic basis of these interventions remains poorly understood. We decompose refusal steering into component-level interventions across four open-weight models, identifying the sparse subsets of attention and MLP components whose steering suffices to reproduce the full behavioral effect. We find that refusal directions concentrate in sparse component mechanisms comprising 28--48\% of upstream components, retaining 88--101\% of steering effectiveness. Within these mechanisms, effective steering further concentrates in approximately 50\% of residual stream dimensions, retaining 85--98\% of the component-…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.06902v1 Announce Type: new Abstract: Retrieval-augmented generation grounds language models in external context, but for long documents flat top-$k$ retrieval can cluster on a single region and miss complementary evidence. RAPTOR-style summary trees address this by recursively clustering chunks and using a language model to summarize each cluster at indexing time, then ranking summary nodes alongside raw chunks at query time. We show the main benefit of summary trees in long-document QA can come from navigation rather than the generated summary content. We introduce NavTree, a leaves-only retriever that builds a deterministic balanced segment tree over chunks (zero language-model calls at indexing) and uses the tree purely as a navigation scaffold: a…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Reka has released Rho-1, a 19B omni-reasoning model trained from scratch. One network reads and generates text, images, video and robot actions over a shared KV cache. A distilled variant returns a 5.3-second clip in about a second. It is a research preview with no public weights yet. The post Reka Releases Rho-1: A 19B Omni-Reasoning Model That Understands, Generates Video and Outputs Robot Actions in One appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Reflection AI has introduced Beam, its first open-weight model. It is a 501B sparse Mixture-of-Experts model with 23B active parameters, built for coding and agentic work. Reflection says it matches GLM-5.2 on reasoning with 3 to 4x less inference compute. Apache 2.0 weights are due later in October 2026. The post Reflection AI Introduces Beam: A 501B Open-Weight MoE Model With 23B Active Parameters for Coding and Agentic Workloads appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Alibaba's Qwen went from an invite-only chatbot in April 2023 to a 2.4-trillion-parameter open-weight model in August 2026. This is the full story, release by release: every major model, its key feature, and how its license changed. Each claim links to its source. The post The Story of Qwen: Alibaba’s AI Models From 7B to 2.4T appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.02444v1 Announce Type: new Abstract: Large language models often solve a theorem forward yet fail to disprove a closely related false one: a falsification gap that supervised fine-tuning does not close and can actively worsen. We frame counterexample generation as constrained witness emission against a deterministic per-theorem Python verifier, and release SymCE, a corpus of 4,707 false undergraduate-algebra and real-analysis conjectures, each paired with executable verifiers. The verifier also serves as the reward function, making SymCE a training environment. Training Qwen3-4B with SFT followed by GRPO under this oracle reveals an imitation trap: counterexample-only SFT collapses true-theorem recognition from 0.27 to 0.00, while RLVR with a sparse…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.02251v1 Announce Type: new Abstract: Speculative sampling accelerates diffusion generation by verifying inexpensive draft states in parallel while preserving the target law. Recent tree-based methods allocate the parallel compute budget more effectively than single-chain drafts, as demonstrated by Diffusion Greedy Rejection Sampling (D-GRS). D-GRS generates $K$ conditionally independent candidates per node, and sequentially tests them in their generation order. Yet the sampled candidates admit an informative ranking without additional target-model evaluations. To exploit this, we introduce Rank-Aware Speculative Sampling (RASS), a verification rule for speculative draft trees based on rank-aware list coupling. RASS orders draft candidates along the p…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.02267v1 Announce Type: new Abstract: Agent harnesses make many small, typed decisions per task: which model to call, which tool to use, whether retrieved text is relevant, whether an input carries an injection. System-1 decision models answer such questions in a single forward pass with class probabilities, promising large cost and latency savings over LLM calls. We present a paired evaluation of an open-weight (Laya) and a hosted (Jev) System-1 model on 11 agent decision points built from 18 public sources: 7,283 base cases plus 6,640 robustness variants, with byte-identical inputs, paired tests, and cross-hardware and cross-day reproducibility checks. Jev is significantly more accurate on 9 of 11 decision points (+10.8 to +46.0 pp). Neither model b…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Cantina Security, with Yeta Labs, has released apex-flash-1, an open-weights model trained specifically for vulnerability research. It is a reinforcement learning fine-tune of Z.ai’s GLM-5.3-Flash, released on Hugging Face under the MIT license. Is it deployable? Yes, the MIT weights serve on vLLM, SGLang or Transformers, but BF16 needs roughly 640 GB of GPU memory. […] The post Can an Open Model Do Security Research? Cantina’s apex-flash-1 Solves 40 of 60 Held-Out Bug Tasks appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Aleph Alpha has released Kolibri, a 78.1B-parameter English-German Mixture-of-Experts model that activates only 3.46B parameters per token. It has a 1M-token context and per-request reasoning effort, and its Apache 2.0 FP8 weights run on a single B200 or H200. The post Aleph Alpha Releases Kolibri: A 78.1B Open-Weight English-German MoE Model With Only 3.46B Active Parameters appeared first on MarkTechPost.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.00053v1 Announce Type: new Abstract: Four-bit floating-point (FP4) Tensor Cores accelerate matrix multiplication, but scale computation, operand packing, layout construction, and saved backward state can erase the gain. We present \emph{format-aware fusion}, which co-designs each quantization producer with its scale domain and consumer layout for native \mxfp{}, global \nvfp{}, and cooperative-thread-array-local \nvfp{}. We evaluate Llama-3-family 8B pretraining through 160 billion tokens using bfloat16 output projections and compiled cross entropy. In matched same-accelerator probes, bfloat16 and Transformer Engine \nvfp{} reach 18.8K and 27.6K tokens/s/GPU, while our fastest custom route reaches 37.9K. \mxfp{} with row-gradient stochastic rounding…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.00049v1 Announce Type: new Abstract: Graphics processing unit (GPU) generations scale matrix, special-function, and memory pipelines at different rates, so kernel bottlenecks move as hardware evolves. FlashAttention-4 exposed this imbalance inside attention on NVIDIA Blackwell. We test whether short polynomial programs can accelerate other special-function-unit (SFU) operations in large language models (LLMs). We first compare native PyTorch evaluation with packed fused multiply--add (FMA) programs in an isolated IEEE binary16 (FP16) sweep spanning L2-resident and high-bandwidth-memory (HBM)-resident working sets. We then replace native sigmoid, tanh, and sigmoid linear unit (SiLU) with degree-3 or degree-4 bfloat16 (BF16) programs in four GB200 inte…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：After leading Meta’s Llama models, Ahmad Al-Dahle is now transforming Airbnb with AI — from how its teams develop products to how it serves guests.
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.00003v1 Announce Type: new Abstract: Vision models pretrained for frame-level appearance often struggle to infer hidden physical properties from motion. We study center-of-mass (CoM) localization for opaque, asymmetric rigid bodies from short monocular videos, where surface cues and point tracking are unreliable under self-occlusion. We propose STATERA, which adapts a pretrained video backbone (V-JEPA) with mostly frozen weights and a lightweight temporal tubelet mixer to predict per-frame CoM heatmaps and trajectories. To support this task, we introduce the HiddenMass Benchmark, comprising 50K MuJoCo trajectories and a 63-sequence real-world test set with physically calibrated CoM ground truth. In simulation, STATERA-50K-Sigma improves normalized Co…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2609.38196v1 Announce Type: new Abstract: Accurate time series forecasting is critical across various domains, yet traditional ensemble methods often suffer from the disproportionate influence of extreme forecasts. We introduce the Conformal Adversarial Generative Ensemble (CAGE), a novel framework that combines generative modeling, adversarial discrimination, and conformal prediction to enhance forecast reliability and accuracy. CAGE employs multiple generative models to produce initial forecasts, which are then evaluated by a discriminative component using conformal prediction techniques. P-values derived from nonconformity scores help dynamically adjust model weights, minimizing the impact of unreliable forecasts. This approach ensures that only the mo…
AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2610.00047v1 Announce Type: new Abstract: Diversity collapse in parallel chain-of-thought has motivated inference-time interventions built on a natural design: when a process reward model (PRM) prunes a chain, its high-PRM prefix is extracted and grafted verbatim as an in-context demonstration into a still-decoding sibling. We isolate this mechanism, PRM-Pruned Fragment Grafting (PPFG), as the most cost-minimal operationalization of cross-trajectory step-level transfer, and test it at the operating point where prior fragment-grafting work reports gains only under additional compensating ingredients. On Qwen2.5-7B-Instruct with Math-Shepherd on full MATH500 (n=500, three seeds), PPFG in both stagnation- and random-targeting variants is statistically indist…