Skip to content
AI NEWSSee every step of AI
Bookmarks
THE JOURNAL
AI NEWS JOURNAL
AI, IN PERSPECTIVE

Search AI news

Search headlines, summaries and articles to connect the dots.

Newest first

llama.cpp releases

b11918

opencl: initial support for q3_k (#31172) * opencl: add q3_k AOS mv * opencl: add q3_k SOA mv and tiled mm * opencl: add q3_k noshuffle gemm/gemv for adreno * opencl: add q3_k dp4a gemm for adreno Co-authored-by: Hongqiang Wang * opencl: tune q3_k kernels for adreno Co-authored-by: Hongqiang Wang --------- Co-authored-by: Hongqiang Wang **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11918/llama-b11918-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11918/llama-b11918-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/ll

Timeline and related coverage
腾讯新闻 · AI频道

After integrating 15 nationally popular apps into Lingguang, I created the most user-friendly domestic Personal Agent.

Since I got my hands on Ant Lingguang 2.0, the beta version of this Personal Agent, it has become my personal butler. To date, I’ve integrated 15 nationally popular apps—via Skills, MCP, GitHub projects, or built-in methods. Basically, whenever I open the “Help Me Manage” section, the scheduled tasks automatically refresh information worth paying attention to. Lingguang isn’t the first Personal Agent I’ve tried; over time, I’ve tested a total of 12. By the later stages, I found myself increasingly…

把15个国民级App接入灵光后，我造出了最舒适的国产Personal Agent
AI腾讯新闻
Timeline and related coverage
SiliconANGLE AI

Cisco Partner Summit: Five insights that will define the channel in the AI era

For most of its 30-year history, Cisco Partner Summit has focused on how partners sell Cisco Systems Inc. technology. This year’s event in Vancouver was about something different: how partners make money in a world where the transaction is no longer the main event. Cisco entered the event with tremendous momentum, as its stock was […] The post Cisco Partner Summit: Five insights that will define the channel in the AI era appeared first on SiliconANGLE.

Timeline and related coverage
腾讯新闻 · AI频道

GPT-6.1 Terra released, with comments: It’s all downloaded.

On the very same day that Anthropic officially filed its IPO prospectus, OpenAI unveiled its new model, GPT-6.1 Terra. This marks the first update to the Terra series in more than three months, and it has already been 27 days since the release of GPT-6.1 Sol on September 29. In its official promotional materials, OpenAI emphasized: “Delivering fast and powerful computing capabilities at a lower price.” However, users don’t seem to be convinced. “Bro, what we need is a model stronger than Opus 5.5. Nobody cares whether it’s faster or cheaper,” one user bluntly remarked, adding that they’ve almost stopped using the Codex subscription because it’s simply not useful anymore. Over the past period, Anthropic has garnered significant attention—particularly in programming-related fields—thanks to Claude Opus 5.5. Many users are now eagerly anticipating Open...

GPT-6.1 Terra发布，评论区：拉完了
AI腾讯新闻
Timeline and related coverage
arXiv cs.CV

ProtoEEG: Structured Prototype Alignment for EEG Visual Decoding

arXiv:2610.33065v1 Announce Type: new Abstract: EEG visual decoding remains difficult because scalp recordings are noisy, spatially coarse, and variable across trials and subjects. Most systems align each trial with a dense image embedding, although that embedding mixes semantic content with instance-specific texture, pose, and background. This target can demand details absent from EEG and provides no explicit account of the visual factors supporting a prediction. We propose \textbf{ProtoEEG}, a prototype-mediated framework that reformulates EEG--image alignment as soft composition over a shared learnable visual codebook. A trainable EEG encoder and a frozen visual teacher produce assignment distributions over the same $K=64$ prototypes. Their weighted reconstructions form a structured rep

Timeline and related coverage
arXiv cs.LG

A Mathematical Theory of the Jacobian Lens

arXiv:2610.31955v1 Announce Type: new Abstract: Recent work from Anthropic introduced a new interpretability tool, called the Jacobian lens (J-lens, Gurnee et al., 2026). This method identified linear directions in the hidden states of frontier language models that had properties reminiscent of working memory-like representations. However, it is unclear what mechanistic properties of such features make them detectable to the J-lens. To this end, we introduce a mathematical theory of the J-lens, and show that it preferentially identifies coherent features -- directions that act like leading, spatially coherent Lyapunov modes. Small perturbations along these directions amplify across layers, remain roughly constant across token positions, and are consistent across different contexts. While w

新闻源配图 · arXiv cs.LG
Timeline and related coverage
arXiv cs.DC

MANTA: Unlocking DAG Flexibility in Asynchronous BFT

arXiv:2610.34419v1 Announce Type: new Abstract: DAG-based BFT protocols marked a shift in consensus design by replacing single-leader proposal paths with concurrent block proposals linked through a shared graph, where references encode consensus evidence. However, under heterogeneous wide-area networks (WANs), existing designs suffer from prolonged rounds and degraded performance due to straggler effects. The root cause is rigid round advancement: replicas move in synchronized rounds, waiting for enough references from the previous round before advancing. We propose a flexible DAG growth framework for asynchronous BFT consensus. Our key observation is that while reference quorums are necessary for safety and liveness, they need not block every round advancement. Instead, the connectivity r

Timeline and related coverage
arXiv cs.LG

PRISM: Promoting Sampling Diversity of Large Language Models with Implicit Stochastic Modulation

arXiv:2610.32101v1 Announce Type: new Abstract: Large language models (LLMs) exhibit mode collapse, where their output distribution concentrates on familiar semantic modes. When diversity disappears, repeated sampling becomes semantic repetition rather than discovery, making rare but stronger solutions unlikely to surface and leaving self-evolution and continual learning difficult to sustain. Most remedies intervene only at inference time, shaping sam- ples without making diversity a model capability. We propose PRISM, a training-time framework that internalizes diversity within the model. PRISM combines (i) submodular curation that pairs each prompt with diverse-yet-correct multi-teacher solutions, and (ii) stochastic embedding modulation, which turns special-token perturbations into impl

Timeline and related coverage
arXiv cs.CL

TraceQuery: learning what to recompute from full-denoising trajectory in diffusion language models

arXiv:2610.33008v1 Announce Type: new Abstract: Diffusion large language models (dLLMs) enable flexible generation through iterative masked denoising, but incur substantial inference cost because token representations are repeatedly recomputed across denoising steps. Existing cache-based methods reduce this cost by reusing intermediate results across denoising steps and recomputing only a subset of positions at each step, chosen either on a fixed schedule or by heuristic signals of change. Our trajectory experiments instead point to decoding trajectory as the more relevant quantity. Cached generation remains accurate when it stays close to the full-denoising trajectory, with errors concentrated around unmask-order deviations and reduced as longer reference prefixes are recovered. This sugg

新闻源配图 · arXiv cs.CL
Timeline and related coverage
arXiv cs.CL

SLICE-Mem: Salience-Gated Lightweight Information Curation for Efficient LLM Agent Memory

arXiv:2610.33015v1 Announce Type: new Abstract: Long-term memory enables large language model (LLM) agents to maintain personalized and temporally coherent interactions, but repeated LLM inference for memory admission, extraction, and organization introduces substantial cost and latency. We present SLICE-Mem, a lightweight memory framework that combines salience-guided admission and predicate-centered information extraction. A compact classifier identifies user turns worth retaining, and a semantic-dependency-based extractor converts admitted turns into concise memory entries. This online pipeline requires no generative LLM calls, while cross-entry reconciliation is deferred to an offline consolidation stage. We evaluate SLICE-Mem on the weekly subset of Memora and LongMemEval-S using 8B a

Timeline and related coverage
arXiv cs.CL

AdaptTok-Indic: Script-Aware Adaptive Tokenization for Indic Languages

arXiv:2610.33017v1 Announce Type: new Abstract: Multilingual tokenizers trained on English-dominated corpora often over-segment Indic languages, resulting in longer token sequences and increased tokenization disparity across scripts. We present AdaptTok-Indic, an Indic-specialized extension of adaptive tokenization that combines Unicode-normalized preprocessing with explicit Indic conjunct handling, script-aware routing, and learned byte-level boundary prediction jointly trained with a small Transformer language model. Evaluated on Bengali and Hindi, with additional evaluation on Gujarati, Tamil, and Telugu, AdaptTok-Indic achieves lower token fertility than SentencePiece BPE, an Indic-oriented tokenizer baseline, and an independently implemented MAGNET-style baseline. Across five recorded

Timeline and related coverage
arXiv cs.CL

Whispering Wisely: Smart Subset Selection for ASR Fine-Tuning

arXiv:2610.32997v1 Announce Type: new Abstract: Automatic Speech Recognition (ASR) models such as OpenAIs Whisper have substantially improved speech recognition quality but remain expensive to fine-tune, especially when large amounts of transcribed audio are available. This makes it challenging for many practitioners to adapt these models to new languages or domains. In this work, we introduce an embedding-based subset selection approach for selecting smaller training sets for ASR fine-tuning. Using Whispers pretrained encoder, we map each audio sample into a fixed-dimensional embedding space and apply a greedy k-center algorithm to select diverse and representative subsets of the training data. We evaluate this approach on both Hebrew and Danish, showing that k-center selection often impr

新闻源配图 · arXiv cs.CL
Timeline and related coverage
arXiv cs.CL

OpenESM: Energy-Steered Models Are Scalable Foundation Learners

arXiv:2610.33005v1 Announce Type: new Abstract: Energy-based models have shown promising properties for domain-specific tasks but remain challenging in modeling general-purpose scenarios. To bridge this gap, we design energy-steered models (ESMs) for foundational language modeling. ESMs integrate energy steering and token generation within a single architecture to leverage the energy-guided search to refine token generation. Specifically, ESMs first optimize a latent state over the energy landscape, then leverages the refined latent state for token generation. By performing such energy-guided pre-training on large-scale corpora, ESMs learn generally effective energy landscape where favorable tokens correspond to lower energy values. To explore the scaling behavior of ESMs, we pre-train ESM

新闻源配图 · arXiv cs.CL
Timeline and related coverage
arXiv cs.CV

A Canary in the Neural Mine: Catching an Image Classifier's Silent, Confident Failures

arXiv:2610.33063v1 Announce Type: new Abstract: Deployed vision models auto-accept confident predictions, so high-confidence errors propagate silently. Maximum Softmax Probability (MSP), the standard failure predictor, has a structural blind spot: as a scalar read from the classifier head, it is first-order blind to feature variation orthogonal to its confidence gradient. Through a factorial analysis of feature-space detectors, we identify a conditional-redundancy principle explaining why they fail to complement MSP: the class-separability geometry they measure is already encoded in MSP and collapses under fusion. The signal that survives is class-conditional typicality on the off-manifold residual---and because this residual's distribution varies by backbone, from near-Gaussian to heavy-t

Timeline and related coverage
arXiv cs.CV

EKT-Router: Expert-Evidence Alignment for Embodied Vision-Language Models

arXiv:2610.33044v1 Announce Type: new Abstract: Embodied vision-language models must answer questions from partial, evolving observations. Their success depends on both the reasoning capability selected and the evidence available to it. Expert routers adapt internal computation, while tool-augmented systems acquire external evidence; choosing these independently can leave a suitable expert without the evidence it needs. We introduce EKT-Router, a bounded framework that jointly selects a capability expert, optional knowledge, and perceptual tools. Outcome-derived signatures represent the overlapping strengths of General, Spatial Reasoning, and Spatiotemporal experts. A joint utility scores complete expert-evidence routes under a call budget. After acquisition, a packet verifier admits valid

新闻源配图 · arXiv cs.CV
Timeline and related coverage
arXiv cs.LG

ReMedI: Inference-Time Refinement and Self-Distillation for Controllable Molecular Optimization

arXiv:2610.32103v1 Announce Type: new Abstract: In drug development, lead optimization must improve competing properties, such as increasing permeability and reducing toxicity, while limiting scaffold changes that could compromise the lead's potency. RL-based post-training, such as C-MORAL, addresses this problem by fine-tuning molecular language models to directly optimize molecular leads for objectives of interest, which relies on expensive task-specific datasets and faces challenges in achieving multi-objective optimization (MOO). We argue a pre-trained model may already contain useful molecular edits that zero-shot generation cannot efficiently reach. Motivated by the limited search capabilities of zero-shot generation, we develop an efficient Metropolis-Hastings search strategy to bet

Timeline and related coverage