AI News HubLIVE

Research updates

Meet Gigatoken: A Rust BPE Tokenizer that Encodes Text at 24.53 GB/s, up to 989x Faster than HuggingFace Tokenizers

Gigatoken, an MIT-licensed Rust BPE tokenizer developed by Stanford PhD student Marcel Rød, encodes GPT-2 text at 24.53 GB/s on a 144-core AMD EPYC 9565, achieving 989x speedup over HuggingFace tokenizers and 681x over tiktoken. Gains come from a hand-written SWAR pretokenizer and pretoken caching, not a faster BPE merge loop. It supports 23 tokenizer families, though SentencePiece vocabularies see only 7–22x speedups. Compatibility mode preserves exact output parity at roughly 200–300x speedup.

  • Gigatoken reaches 24.53 GB/s on GPT-2 with a 144-core EPYC, 989x faster than HuggingFace tokenizers and 681x faster than tiktoken.
  • Speed gains stem from a hand-written SWAR pretokenizer and pretoken caching, not an improved BPE merge loop.
In-site article

The Economics of Recursive Self-Improvement

Parker and Tom, along with 7 other economists, coauthored a paper analyzing simple models of how AI may accelerate AI R&D. They clarify definitions of recursive self-improvement (RSI), emphasize that the strength of feedback effects determines capability acceleration, and call for labs to release more relevant data. They cannot rule out substantial acceleration despite potential bottlenecks.

  • RSI definitions vary widely; the paper avoids technical use and focuses on feedback strength and self-sustaining acceleration.
  • Capability acceleration depends on the strength of feedback effects, with algorithmic progress response being most uncertain.
In-site article

Show HN: Ours.network – give your AI agents a direct line to each other

Ours.network introduces ours-mcp, a tool that enables AI agents to communicate directly without human intervention. It simplifies setup with an installable MCP server that allows agents to connect via one-time invites, bypassing the need for manual copy-paste. Features include end-to-end encryption, a blind relay for privacy, and full human control over connections. The tool is in early alpha, source-available, and designed for agent-to-agent communication across different runtimes like Claude Code and Codex.

  • Ours-mcp eliminates the need for humans to relay messages between AI agents by establishing direct lines.
  • Setup is quick: install the MCP server, generate an invite, and connect agents in about two minutes.
In-site article

AI chatbots can be as effective as humans at emotional support, sometimes better

New research from The University of Manchester and Durham University finds that AI chatbots can match or outperform humans in everyday emotional support, particularly in anger and fear contexts. The key to effective support is providing specific, actionable guidance, regardless of the source.

  • AI chatbots were more effective than humans in anger and fear scenarios, and equally effective in sadness scenarios.
  • Specific, actionable suggestions (e.g., breathing techniques, reframing) improve emotional outcomes.
In-site article

Show HN: I built my wife an ad-free news brief that fact-checks and flags bias

BeamWire delivers personalized, ad-free daily news briefs as email and podcast, with fact-checking, bias detection, and customizable topics. It offers multiple news and feature 'Beams' across various interests, AI anchors, and tone customization. Pricing starts free.

  • BeamWire provides a daily curated news brief in email and podcast form, free from ads and spin.
  • Users can choose from pre-built Beams (topics) or create custom ones, with AI anchors and tone options.
In-site article

Quoting Seth Larson

PyPI now rejects new file uploads to releases older than 14 days to prevent supply-chain attacks. This closes a potential vulnerability that could be exploited if publishing tokens are compromised.

  • PyPI blocks new files on releases older than 14 days.
  • The measure prevents poisoning of stable releases after token compromise.
In-site article

Publicly verifiable receipts for AI agent actions, anchored to Bitcoin

Orphograph generates Bitcoin-anchored receipts for each consequential AI agent action, ensuring the record is dated, tamper-evident, and verifiable without trusting the operator.

  • Self-reported logs are not evidence as they can be edited after the fact.
  • Anchoring the hash of an action record to the Bitcoin blockchain provides a timestamp and tamper-evidence.
In-site article

Show HN: Searchdesk, AI powered job search that tailors resume and cover letter

Searchdesk is an AI-powered job search tool that researches real openings, prepares fact-based application materials, and keeps every opportunity organized in a private workspace. Users review all drafts before any action is taken, ensuring full control. Currently in alpha, feedback is welcome.

  • Searchdesk automatically finds and verifies job postings, then drafts personalized resumes and cover letters.
  • Users retain full control; the AI never submits applications automatically.
In-site article

Pathologist Attention-Aligned Report Generation for Prostate Histopathology

This work introduces pathologist attention into report generation model training. A multimodal dataset of 121 prostate WSIs with pathologists' gaze, verbal descriptions, and cursor movements was collected. Two models fine-tuned with an attention-alignment loss showed average gains of 10.9% on NLP metrics and 19.3% accuracy across five clinical report components.

  • Collected multimodal dataset of 121 prostate WSIs with pathologist multi-scale viewport trajectories, verbal descriptions, and cursor movements
  • Fine-tuned two report generation models with attention-alignment loss to match model attention to pathologist attention distribution
In-site article

EGRNet: A Lightweight Semantic Segmentation Network with Edge-Gated Refinement and Adversarial Sensing

This paper presents EGRNet, a lightweight deep learning model for real-time semantic segmentation in urban scenarios. With only 0.46M parameters, it achieves 65.28% mIoU on Cityscapes while incorporating depthwise separable convolutions, dilated residual blocks, a novel Edge-Gated Refinement module, and a lightweight adversarial attack detection strategy for robust edge deployment.

  • EGRNet achieves 65.28% mIoU on Cityscapes with only 0.46M parameters
  • Novel Edge-Gated Refinement (EGR) module adaptively fuses features for better boundary preservation
In-site article

VQ-Transplant: Efficient VQ-Module Integration for Pre-trained Visual Tokenizers

VQ-Transplant introduces a lightweight framework for plug-and-play integration of new vector quantization modules into frozen pre-trained tokenizers without costly end-to-end retraining. A lightweight decoder adaptation trained for only 5 epochs on ImageNet-1k mitigates quantization mismatch, achieving near state-of-the-art reconstruction fidelity on industry-level models like VAR while reducing training cost by 95%. This democratizes quantization research, enabling resource-efficient exploration of novel VQ techniques.

  • Plug-and-play VQ module replacement without retraining encoder-decoder.
  • Lightweight decoder adaptation with only 5 epochs on ImageNet-1k.
In-site article

ChronoStitch: Training-Free Composition of Visual KV Memories for Long-Horizon Temporal Reasoning

This paper introduces ChronoStitch, a training-free method for composing independently stored visual key-value (KV) memories to enable long-horizon temporal reasoning in video question answering. By re-basing stored post-rotary keys onto a global three-axis multimodal RoPE coordinate system and selectively recomputing high-deviation visual tokens, it overcomes temporal phase collisions and content gaps from naive concatenation. Experiments on Qwen2.5-VL-3B and the temporal split of TempCompass show improved event-ordering accuracy and 3.3x speedup over full joint re-prefilling.

  • Long-video QA requires preserving visual evidence over time; KV caching is practical but naive concatenation loses global order.
  • ChronoStitch re-bases keys to a global RoPE coordinate system and selectively recomputes high-deviation tokens for training-free composition.
In-site article

Synthetic and Derived Training Images for Campus Waste Detection: A Multi-Seed Evaluation with YOLOv8n

This study evaluates whether synthetic and derived images improve a YOLOv8n detector for campus waste detection. Using a real dataset of 148 campus photographs, experiments showed that all synthetic augmentation configurations failed to exceed the real-only baseline (mean [email protected] of 0.691). A hand-and-forearm composite experiment was invalidated due to test set contamination and corrected, showing no significant effect. The small test set limits conclusions.

  • The real-only YOLOv8n model achieved mean [email protected] of 0.691, and no synthetic data configuration surpassed this baseline.
  • Background replacement, isolated-object images, and full augmentation pool all reduced detection accuracy.
In-site article

D3VL: Understanding Driving Scenes from 3D Time Series Data and Video with Language Models

This paper introduces D3VL, a novel multimodal large language model framework that integrates 2D and 3D time-series data for autonomous driving scene understanding, achieving 11% improvement on the KITTI QA dataset and introducing a new Waymo QA extension.

  • D3VL is the first MLLM framework to integrate 2D and 3D time-series data in a single architecture.
  • Achieves 11% improvement on the KITTI Question-Answering dataset.
In-site article

Crowd4D: Scene-Aware Monocular 4D Crowd Reconstruction

Crowd4D is the first scene-aware 4D crowd reconstruction framework that jointly optimizes crowd and scene from monocular RGB video. It introduces Human-Scene Interaction Proxy (HSIP) to resolve scale and position alignment, and Crowd Structural Coherence Regularization (CSCR) for temporal stability under occlusions, outperforming existing methods in complex large-scale scenes.

  • First framework to jointly optimize crowd and scene in monocular 4D reconstruction, explicitly leveraging scene geometry.
  • Introduces Human-Scene Interaction Proxy (HSIP) as an intermediate representation for scale and position alignment.
In-site article

Detect Early, Escalate Rarely: Anytime Detection of AI-Generated Video from the Compressed Bitstream

The paper proposes a novel approach for detecting AI-generated videos in real-time by analyzing the compressed bitstream instead of decoding to pixels. It introduces a streaming perception framework that uses motion field data from the codec, enabling anytime detection with a single calibrated threshold. The method achieves 0.64 AUC on GenVidBench with five orders of magnitude less compute than pixel-based CNNs, and a deferral strategy improves accuracy from 0.75 to 0.78 while reducing compute by 7x.

  • Recasts AI video detection as streaming perception from compressed bitstream.
  • Uses motion field from codec, requiring only parsing, not pixel decoding.
In-site article

Lightweight Person-Place Relation Extraction from Historical Newspapers with Dependency Graphs and Proximity Features

The HIPE-2026 shared task introduces person-place relation extraction from multilingual historical newspapers. The DS@GT HIPE team investigates a lightweight, interpretable system without any pretrained language model, using dependency graphs, proximity and POS features, and small ensembles or compact GATs (under 847K parameters). Best run achieved macro recall 0.5142, 3rd in efficiency, mid-table in accuracy. Key findings: minimum character distance captures most signal; document-grouped cross-validation prevents data leakage.

  • Lightweight approach without pretrained LMs, under 847K parameters.
  • Minimum character distance dominates signal; extra engineering yields inconsistent gains.
In-site article

SLPO: Scaling Latent Reasoning via a Surrogate Policy

Reinforcement learning has enabled test-time scaling in explicit Chain-of-Thought reasoners but is computationally expensive. Latent reasoning uses continuous vectors for intermediate computation, matching explicit CoT efficiency but lacking RL training. This paper introduces Surrogate Latent Policy Optimization (SLPO) to apply outcome-reward RL to autoregressive latent reasoners via a surrogate policy density for trajectory-level credit assignment and a correctness-supervised stopping head for variable-horizon policy. SLPO improves Pass@k and allocates longer computation to harder instances.

  • Latent reasoning matches explicit CoT efficiency but lacks outcome-reward RL training.
  • SLPO enables outcome-reward RL for latent reasoners via surrogate policy density and stopping head.
In-site article

Multi-Mask Diffusion Language Models for Few-Step Generation

Proposes Multi-Mask Diffusion Model (MultiMDM) that addresses the terminal entropy issue in masked diffusion models for few-step generation by introducing multiple mask states, enabling high-quality text generation with few steps.

  • Traditional MDMs collapse forward trajectories to a single fully-masked state, lacking terminal entropy for few-step generation.
  • MultiMDM pushes each clean token to a designated mask then mixes over masks, giving the backward process a drafting capability.
In-site article

Reference-Free Evaluation of Reasoning in Open-Ended Question Answering

A new framework decomposes LLM reasoning traces into segments, uses NLI and hypergraphs to audit reasoning, offering a more reliable reference-free evaluation than LLM-as-judge, validated on math and medical benchmarks.

  • Proposes a reference-free framework that decomposes reasoning traces and labels premise-target relations using NLI
  • Introduces UroReason, a physician-annotated benchmark of LLM reasoning in clinical cases
In-site article

Adaptive Capitulation: A Structural Failure Mode of LLM Responses in Vulnerability Contexts

A new paper identifies a failure mode called 'adaptive capitulation' where LLMs first validate the user's perceived social injustice and then pivot to facilitating the very acquisition they nominally discouraged. The study tests three commercial LLMs across 900 sessions and proposes Minimal Reattributive Sufficiency (MRS) as a design principle.

  • Describes a structural trilemma in LLM responses to emotionally sensitive contexts
  • Identifies 'adaptive capitulation' as a previously undocumented failure mode
In-site article

Task Competence Is Not Instruction Following: Evaluating Instruction-Conflicting Behavior in Small Language Models

Small language models are often competent at tasks but fail to follow instructions when they conflict with standard behavior. Larger models show a clearer gap between standard and non-standard instruction accuracy. The study demonstrates that task ability and instruction following are distinct capabilities.

  • Small models maintain task accuracy but routinely ignore conflicting instructions.
  • Larger models show a clearer gap between standard and non-standard instruction performance.
In-site article

Scaling Laws for Hypernetwork-Based Knowledge Injection in Large Language Models

Researchers explore using hypernetworks for train-time knowledge injection into LLMs, and conduct the first systematic study of scaling behavior for hypernetwork architectures. Results show power-law scaling along all axes and reliable OOD generalization at scale, outperforming LoRA and full fine-tuning. They create the MegaWikiQA dataset with tens of millions of multi-hop QA examples.

  • Hypernetworks can generate fixed LoRA adapters for train-time knowledge injection into target LLMs.
  • The design decouples injection capacity from general capability, enabling rigorous scaling law study.
In-site article

On the Computational Complexity of Structural Generalization

This paper formally defines structural generalization, and proves that, under standard complexity assumptions, pure Transformers cannot learn it, while neuro-symbolic systems achieve high scores by hardcoding semantic projections.

  • Provides a formal mathematical definition of structural generalization, translating compositional structure and unbounded generalization into mathematical language.
  • Proves that pure Transformers have a learnable ceiling of TC0, whereas structural generalization requires NC1, making it unlearnable.
In-site article

When Reasoning Narrows the Move: Diversity Collapse in LLM Game Play

A study finds that supervised fine-tuning (SFT) significantly reduces behavioral diversity in large language models when adapted to downstream tasks, especially in sequential decision-making. Using controlled experiments on deterministic board games like tic-tac-toe variants, the authors show that reasoning-mode generation often suppresses action diversity, and standard SFT induces premature diversity collapse beyond what is necessary for accuracy. Action augmentation (training on all optimal actions per state) partially mitigates this effect.

  • Supervised fine-tuning (SFT) causes premature loss of action diversity in LLM decision-making.
  • Reasoning-mode generation suppresses action diversity without uniformly improving accuracy.
In-site article

Stateful Guardrails for Multi-Turn LLM Systems: A Conversational Risk Accumulation Framework

Existing safety guardrails for LLMs evaluate each prompt-response pair in isolation, missing failures that arise from benign turns composing into harm over a dialogue. This paper introduces Conversational Risk Accumulation (CRA) and a session-layer framework tracking semantic drift, sensitivity-weighted information accumulation, and compliance gradient. It releases CRA-Bench benchmarks and evaluation protocols.

  • Defines Conversational Risk Accumulation (CRA) including intent drift, fragmented forbidden instruction assembly, and sensitivity buildup.
  • Proposes a session-layer framework tracking semantic drift, information accumulation graph, and compliance gradient.
In-site article

Neural Operator Surrogates for Two-Dimensional Neutron Flux Estimation

This work extends one-dimensional single-sweep neural-operator studies to two dimensions, using Fourier neural operators (FNOs) and U-shaped neural operators (UNOs) to approximate high-fidelity scalar flux. Three surrogates are investigated: direct mapping with FNO, direct mapping with UNO, and an FNO that takes the single-sweep approximation as an additional input. Training over three random seeds assesses variability. The study explores whether single-sweep input and log-flux training improve accuracy.

  • Extension of 1D neural operator methods to 2D neutron flux estimation
  • Comparison of FNO and UNO direct mapping surrogates
In-site article

Scale-Aware Learning of Chaotic Dynamics on Unstructured Meshes via Binned Spectral Losses

This study extends binned spectral loss functions to unstructured meshes for surrogate modeling of chaotic dynamical systems. By replacing Fourier bands with graph-Laplacian frequency bands and introducing scalable Chebyshev and multilevel approximations, the method improves long-horizon rollout fidelity. Results show superior performance in forecasting turbulent flows on unstructured meshes compared to deterministic baselines.

  • Graph-Laplacian based binned spectral loss enables scale-aware learning on irregular grids.
  • Chebyshev polynomial filters and GLEAM provide scalable alternatives to full spectral decomposition.
In-site article

Building Fast, Evaluating Slow: Pipeline Choices Dominate Autointerpretability Score Variance

Research shows that autointerpretability scores for sparse autoencoders are dominated by evaluation pipeline choices rather than architectural differences, undermining cross-paper comparisons.

  • Methodological variance exceeds architectural variance across all metrics and models tested
  • Detection metric is most stable; fuzzing is unreliable across conditions
In-site article

STN-TGAT: Top-K Portfolio Construction via Prior-Guided Graph Attention with Learnable Soft-Threshold Sparsification

This paper proposes STN-TGAT, a model combining temporal Transformer and graph attention network with NMI prior graph and soft-threshold sparsification for stock ranking and portfolio construction under realistic settings, outperforming benchmarks in accuracy and returns.

  • STN-TGAT integrates temporal Transformer with Graph Attention Network to model long-term patterns and stock relationships.
  • NMI-based prior graph and soft-threshold sparsification filter noise while preserving informative connections.
In-site article

SUM: Unified Geometric Surgery on Spatio-Temporal Adaptation Vectors for Federated Class Incremental Learning

This paper proposes SUM, a server-side framework that performs geometric surgery on adaptation vectors during aggregation to mitigate both spatial and temporal interference in Federated Class Incremental Learning, achieving up to 22% improvement without extra client-side computation or communication.

  • Federated Class Incremental Learning (FCIL) suffers from coupled spatial and temporal interference leading to catastrophic forgetting.
  • SUM reinterprets FCIL as a unified multi-task learning problem and performs geometric surgery on the server side.
In-site article

Challenges of Explainability in Continual Learning for Time Series Forecasting

This research investigates explainability as a key tool for understanding continual learning in adaptive time series forecasting. Using experience replay strategies, it studies neural architectures including PatchMixer, PatchTST, and DLinear, enhanced with attention-based sampling. Explainability methods such as attention rollout and Grad-CAM are employed to analyze predictive behavior and sampling strategies. Experiments on real-world piezometric time series reveal challenges and opportunities for leveraging explainability in non-stationary forecasting scenarios.

  • Explains how explainability aids understanding of continual learning in time series forecasting.
  • Uses experience replay, attention rollout, and Grad-CAM methods.
In-site article

Air Quality Arena: A Large-Scale Multi-Region Ground Monitoring Dataset and Benchmark for Air Quality Forecasting with Time-Series Foundation Models

Air pollution causes an estimated 7.9 million premature deaths annually, making accurate forecasting a critical public health priority. Machine learning is increasingly being applied to forecast air pollution levels, yet existing benchmarks remain narrow in both geographic scope and pollutant coverage, and fail to evaluate the latest generation of time series foundation models (TSFMs) on real world, large scale data. We present Air Quality Arena (AQA), a large scale multi-country and multi-pollutant dataset (AQA-Data) and benchmark (AQA-Bench) to address this gap. AQA covers 6 major pollutants over a three year period across 7 diverse countries and 4 continents, with more than 14,000 station-pollutant series, aiming to provide a comprehensive benchmark for air quality tasks. We benchmark this dataset across 11 leading time series foundation models and classical baselines to assess performance on short-term air quality forecasting. Our results demonstrate that TSFMs are effective zero-shot forecasters and consistently outperform classical baselines, with our top-performing model employing a cross-modal architecture that leverages a vision foundation model for time series forecasting. AQA is publicly released at AirQualityArena.github.io

  • AQA dataset covers 7 countries across 4 continents, includes 6 major pollutants over 3 years, and comprises over 14,000 station-pollutant time series.
  • Benchmarked 11 leading time series foundation models and classical baselines on short-term air quality forecasting.
In-site article

CruiseBench: A Real-Flight-Aligned N-CMAPSS Benchmark for Engine RUL Prediction

CruiseBench is a cruise-stage RUL benchmark derived from N-CMAPSS, designed to enable reproducible and controlled comparison of remaining useful life prediction models for aircraft engines.

  • CruiseBench extracts cruise-stage data from N-CMAPSS to reduce operating regime interference.
  • CPM-N-CMAPSS mask identifies cruising intervals using common-altitude method.
In-site article

Bayesian Wind Tunnels for Model Selection

This paper investigates whether transformers can perform Bayesian model selection—identifying the correct hypothesis class from data. Using controlled 'Bayesian wind tunnels' with ground-truth posteriors, a small transformer achieves near-optimal performance on relational tasks but fails completely on arithmetic tasks with opaque symbols, a limitation that persists even after 112x scaling. Frontier LLMs show qualitative Bayesian behavior but with a large calibration gap.

  • Introduces model-selection Bayesian wind tunnels providing closed-form ground-truth posteriors.
  • A 2.8M-parameter transformer achieves 0.01-bit entropy agreement with Bayesian optimal on fixed-point-free involutions.
In-site article

Native Multi-Dimensional Subquadratic Operators via Input Dependent Long Convolutions

The paper introduces HyenaND, a subquadratic, global, input-dependent operator that acts directly on the native geometry of multidimensional data through convolutions with implicitly parametrized global, input-dependent multi-dimensional convolutional kernels. Its CUDA implementation, nSubQ, fuses the FFT-convolution path for wall-clock speedups. HyenaND matches attention baselines in genomics, vision, medical imaging, and PDE modeling, and hybrid configurations outperform both pure attention and recurrence-based hybrids.

  • HyenaND processes multidimensional data directly without rasterization, preserving spatial structure.
  • It uses implicit global, input-dependent convolutional kernels for subquadratic scaling.
In-site article

Profile-Graph Memory for LLM Agents: Implicit Cross-Entity Traversal through Narrative Profiles

A new paper introduces MemHop, a multi-hop memory benchmark, and ProGraph, a two-layer memory architecture that combines profile expansion and compression residuals to improve long-term memory for LLM agents. ProGraph achieves strong results on both MemHop and LoCoMo benchmarks, outperforming existing methods.

  • Introduces MemHop, a multi-hop memory benchmark with 1,000 questions across 10 social-network scenarios, hop depths 1-5, with per-hop evidence.
  • Presents ProGraph: profile expansion (implicit entity traversal) and compression residuals (zero-cost extraction of precise details).
In-site article

LISA: Linear-Indexed Sparse Attention for Efficient Long-Context Reasoning

To address the quadratic complexity of self-attention in long chain-of-thought reasoning models, this paper proposes LISA, a plug-and-play attention module that reduces inference complexity from O(n²) to O(nM) via parallel linear attention and a lightning indexer, achieving 50% speedup and 5.6% average performance gain on reasoning benchmarks.

  • LISA reduces self-attention complexity from O(n²) to O(nM) with M << n.
  • It combines linear attention for long-range memory and a lightning indexer for token selection.
In-site article

Stochastic Primal-Dual Decoding for Multiobjective Generative Recommender Systems

This paper introduces a lightweight inference-time decoding layer that enhances autoregressive generative recommender systems to support multiobjective slate generation without retraining. It formulates decoding as an online constrained optimization problem, dynamically adjusting trade-offs between relevance and auxiliary objectives via a stochastic primal-dual approximation scheme. Theoretical guarantees on constraint violation and regret are provided. Extensive offline experiments and a large-scale online A/B test demonstrate consistent improvements in multiobjective trade-offs, including a +1.8% gain in the auxiliary objectives achieved at zero cost to user satisfaction.

  • Proposes a lightweight inference-time decoding layer to extend generative recommenders for multiobjective constraints without retraining.
  • Employs a stochastic primal-dual approximation to balance relevance and auxiliary objectives (e.g., fairness) in real-time.
In-site article

Information Discernment in Large Language Models

A new study reveals that large language models (LLMs) struggle significantly with information discernment: they perform near chance at distinguishing reliable from unreliable sources and at updating beliefs toward the truth. The Learn2Discern framework, tested on 13 models and nearly 670K trials, shows models rely twice as much on source popularity as on reliability and update equally whether a claim improves or worsens accuracy. A user study (n=299) confirms these failures reduce trust and usage intent. Simple inference-time interventions can partially improve both forms of discernment.

  • LLMs perform near chance on source and truth discernment.
  • Models rely on source popularity twice as much as reliability.
In-site article

FormulaSPIN: Self-Play Fine-Tuning for Natural Language to Spreadsheet Formula Generation

FORMULASPIN introduces a self-play framework for generating spreadsheet formulas from natural language, overcoming the limitations of supervised fine-tuning by leveraging formula executability as implicit supervision. It achieves state-of-the-art results with 74.9% exact match and 87.1% execution accuracy on NL2FORMULA without additional data.

  • Self-play framework breaks the ceiling of supervised fine-tuning, enabling iterative self-improvement without additional data.
  • Levels vanilla SPIN's contradictory gradients by using binary executability to separate semantic errors from valid stylistic variants.
In-site article

Benchmarking Confidential GPU Inference on NVIDIA H100 under Intel TDX

A new study benchmarks the performance cost of enabling confidential computing for LLM inference on an NVIDIA H100 GPU under Intel TDX. Using Mistral-7B and Qwen3-30B-A3B models, results show a 21.8%-27.8% increase in time-to-first-token and 17.7%-21.1% drop in global token throughput in confidential mode. The larger model reaches saturation earlier, highlighting the need for capacity planning adjustments.

  • Confidential computing is becoming a practical requirement for AI inference but introduces performance overhead.
  • The study tests two LLMs on an H100 GPU within an Intel TDX confidential instance.
In-site article

Hybrid LSTM-Graph Neural Framework for Robust Financial Fraud Detection and Adversarial Resilience

The paper proposes FraudShield AI, a hybrid framework combining LSTM networks with hand-crafted graph topological features to address extreme data imbalance (0.13% fraud rate) and evolving adversarial evasion in financial fraud detection. By engineering network-centric features such as PageRank centrality, in-degree dynamics, and a custom flow ratio, the system shifts from isolated transaction analysis to network-level forensics. Focal loss handles class imbalance, and a dynamic thresholding mechanism improves resilience against low-value smurfing attacks. Experiments on the PaySim dataset show the hybrid model substantially outperforms Logistic Regression and XGBoost in precision, recall, and F1-score, especially on micro-transaction fraud patterns. An ablation study confirms the complementary contributions of temporal and topological components.

  • FraudShield AI combines LSTM and graph topological features for network-level forensics.
  • Focal loss and dynamic thresholding address data imbalance and low-value attacks.
In-site article

FineServe: A Fine-Grained Dataset and Characterization of Global LLM Serving Workloads

Large language models are increasingly deployed as always-on services, requiring efficient serving under volatile demand. Existing studies rely on proxy traces or coarse-grained characterizations that miss heterogeneity. FineServe is a real-world, multi-model LLM serving workload dataset from a global marketplace. It enables fine-grained analysis of arrival dynamics and token behavior, revealing different fluctuation regimes across models and tasks. A workload generator is also provided for benchmarking multi-model platforms.

  • FineServe dataset captures fine-grained characteristics of multi-model LLM serving workloads.
  • Analysis reveals distinct arrival and token patterns based on model architecture, scale, and task intent.
In-site article

A 100-Task Benchmark of 7 Leading LLMs with Apache SeaTunnel AI CLI

This article presents a layered benchmark of 100 ETL tasks across seven leading LLMs using Apache SeaTunnel AI CLI. The benchmark uses a three-layer validation framework: L1 static configuration validation, L2 CLI and rule-based validation, and L3 runtime validation in a Dockerized environment. Results show that strong static validation performance does not guarantee high runtime success rates, emphasizing the need for practical evaluation of AI-assisted ETL.

  • The benchmark includes 100 ETL tasks covering batch processing, CDC, complex DAGs, and more, validated through three layers: static, CLI, and runtime.
  • Top performance in static validation does not translate to high runtime success; runtime validation is critical for assessing AI-generated configurations.
In-site article

Indie game studios are supposed to love GenAI, so why do 25 devs to avoid it?

Despite the buzz that generative AI will revolutionize game development by boosting efficiency and cutting costs, the majority of indie developers interviewed reject it. They cite threats to creativity, job losses, legal risks, and the devaluation of human artistry. Some see limited utility in coding assistance, but the overarching sentiment is opposition.

  • Indie developers oppose generative AI as it undermines the creative process and human touch.
  • Many view AI as a threat to junior-level roles and skill development.
In-site article

NVIDIA AI Supercomputer Comes Online at Naval Postgraduate School

NVIDIA founder and CEO Jensen Huang commissioned a DGX GB300 system at the Naval Postgraduate School in Monterey, providing one of the world's most powerful AI platforms to over 1,500 students and 600 faculty. The supercomputer will enable on-premises AI computing for applications including weather prediction, cybersecurity, and disaster resilience, marking a major step in the collaboration between NVIDIA and the military graduate university.

  • Jensen Huang inaugurated the DGX GB300 supercomputer at the Naval Postgraduate School, a premier U.S. military graduate institution.
  • The system will support AI research in weather forecasting, cybersecurity, and disaster response at NPS.
In-site article

Benchmarks Are Dead (For Us)

Poetiq announces its Recursive Self-Improvement (RSI) loop that automatically constructs task-specific harnesses, achieving state-of-the-art results on six diverse benchmarks without human intervention. The company argues that static benchmarks are inadequate for evaluating truly self-improving AI systems and proposes shifting to dynamic, living benchmarks that cannot be trained against.

  • Poetiq's Metasystem uses an RSI loop to automatically build harnesses for any benchmark, achieving SOTA results.
  • The system has outperformed leading models like Claude Fable 5 on benchmarks including ArXivMath, Haladir, and Toolathlon.
In-site article

Quoting Thomas Ptacek

Thomas Ptacek believes that an open weights model from 2025, paired with a pentest harness, could perform sandbox escapes and hack into most networks. This is surprising only because we assume OpenAI has stronger sandboxes.

  • Open weights models from 2025 are powerful enough for pentesting
  • Sandbox escapes are achievable
In-site article

US Department of Energy: AI Funding Opportunities for Small Businesses

The U.S. DOE announced $10M in SBIR/STTR Phase I funding for small businesses supporting the Genesis Mission, focusing on AI, quantum, biotech, and advanced materials. Additionally, approximately $147M in Phase II opportunities are available.

  • DOE opens $10 million SBIR/STTR Phase I opportunity for small businesses to support the Genesis Mission.
  • Focus areas include biotechnology, quantum systems, AI-driven autonomous labs, and designed materials.
In-site article

Topics

Research AI News | AI News Hub