spec-driven agentic development reshapes software delivery
a new report formalizes spec-driven agentic development, where machine-readable specs fuel autonomous coding agents across the software lifecycle.
researchtag ai
summaries filed under ai
deepmind partners with eve online studio for ai research
google deepmind and fenris creations are collaborating to build ai agents that can learn, remember, and plan in the persistent world of eve online.
researchinherent's faraday agent beats larger models at replicating research
london startup inherent says its small ai agent faraday outperformed anthropic and openai models at reproducing scientific paper findings.
industryConstraint-Aware GPU Allocator Boosts Utilization by 33 Points
A new GPU scheduler from Dharma-AI improves cluster utilization and priority-weighted output by planning allocations across the entire scheduling horizon instead of using FIFO order.
machine learningBenchmark optimization found in top speech recognition models
New probes show leading ASR models reproduce benchmark transcripts even when audio contradicts them, inflating scores.
machine learningGoogle adds mobility data to language models for better place understanding
Google Research introduces ME-POIs, a framework that blends text and mobility data to improve predictions about places like opening hours and busyness.
researchGoogle's Biomarker Discovery Framework prioritizes wearable sensor candidates
Google Research introduces a multi-agent AI system that structures biomarker discovery from wearable data through iterative hypothesis generation, statistical analysis, and literature-grounded reasoning.
researchAgent memory dosage depends on model capability
IBM Research shows that the right amount of agentic memory varies by model tier, with curated retrieval boosting weaker models and full guideline sets helping stronger ones.
machine learningLiquid AI Releases DSpark Draft Models for Faster LFM2.5 Inference
Liquid AI published DSpark draft checkpoints for three LFM2.5 models, enabling speculative decoding that speeds up inference up to 3.2x on GPU and 2.87x on-device without changing output quality.
machine learningModel Cards Fall Short for Open-Weight Model Governance
An analysis of 500 Hugging Face model cards finds current transparency artifacts insufficient for downstream safety governance of open-weight foundation models.
researchLiquid AI Releases QAD Q4_0 Checkpoints for LFM2.5 Models
Liquid AI releases quantization-aware distillation checkpoints that recover 97% of BF16 accuracy for LFM2.5 models at Q4_0 memory and speed.
machine learningMojo programming language goes open source
Mojo, a Python-inspired language for GPU programming, has released its compiler and toolchain under Apache 2 license.
toolsSentence Transformers v6.0 Adds Multi-Vector Encoder
Sentence Transformers v6.0 introduces MultiVectorEncoder for ColBERT-style late interaction retrieval, supporting PyLate, Stanford-NLP ColBERT, and ColPali checkpoints.
machine learningGoogle's PhotoScan estimates body composition from smartphone photos
Google Research demonstrates a deep learning method that predicts insulin resistance from smartphone imagery with accuracy close to DXA scans.
researchRubricForge induces reward-free judging rubrics for agent evaluation
RubricForge evolves a human-readable rubric from labeled trajectories to reduce over-crediting in language-model agent evaluation.
researchA Year in LLM Serving: Workload Evolution, Caching and Load-Balancing
A one-year production trace from Chutes reveals how LLM serving workloads evolve, with implications for caching and load-balancing.
researchWeekly brief: 2026-08-10-2026-08-16
This week in AI: watermarking for Claude, open model shifts, reasoning debates, and new benchmarks for integrity and factuality.
data scienceAnthropic details how Claude's new watermarks will work
Anthropic explains its text watermarking for Claude, using Google DeepMind's SynthID-Text to comply with the EU AI Act, and addresses editing, code, and detection.
industryFP8 Training on AMD GPUs Now Upstreamed in PyTorch
AMD and Meta engineers upstreamed FP8 training optimizations for AMD Instinct GPUs into TorchAO and TorchTitan, delivering up to 13.4% throughput gains on dense models and recovering 89% of quantization overhead on MoE models.
machine learningReasoning Is a Learnable Rule-Based Process
A position paper argues that AI reasoning lacks clear definitions and proposes treating it as a learnable rule-based process with a best-practices checklist.
research