Category
3,244 papers
Global stable rankings can still leave individual near-tie orderings sensitive to benchmark composition, and sub-one-point leaderboard gaps should be accompanied by evidence that the implied ordering is composition-robust.
Post-training quantization (PTQ) is widely used to reduce the cost of serving large language models (LLMs), but its accuracy cost is uneven and is often tuned per model. We study where quantization damage occurs and how to allocate a small additional precision budget. Using causal mixed-precision intervention as ground...
CRAFT, which stands for Cold-start Reasoning Alignment Alignment via Fine-Tuning, is proposed, a data-centric method to autonomously generate a verified dataset of (input, trace, label) triplets, requiring substantially less compute and wall-clock time than GRPO-based methods.
CrisBench, a 1,200-question benchmark of lifecycle reasoning that combines a controlled formal setting with programs executed against Cordis, a runtime that manages component dependencies and cleanup, and asks models to identify affected components, predict state after a specified teardown order, and determine which co...
Reach audiences
Advertise in front of researchers, engineers, and readers.
A framework is proposed that identifies which label pairs the model struggles to distinguish, expands the candidate set to include confusable labels, and generates targeted rules to differentiate between similar candidates, which requires no fine-tuning and transfers to smaller, cheaper models.
LatentPress is introduced, which writes conversational histories and long documents into a third representation: continuous memory tokens that a frozen decoder reads directly through its input-embedding interface, with no text reconstruction at inference.
This work introduces Power-Law Entropy Search (PLES), a computational cost-aware acquisition function built on multi-fidelity Bayesian optimization that efficiently estimates optimal hyperparameter scaling laws through adaptive experimentation using less than one-tenth of the computational budget required by convention...
DronesCATS-Agent, an architecture where the MLLM is a swappable component, and DroneCATS, a benchmark treating the model as the independent variable are introduced, an architecture where the MLLM is a swappable component and DroneCATS, a benchmark treating the model as the independent variable.
Inspicio is presented, an open-vocabulary retrieval pipeline that links tokens in context to synsets of the Open English WordNet (McCrae et al., 2020) without requiring any source-language inventory or mapping.
This work compares nine post-training arms spanning GRPO, SFT, and DPO from one starting checkpoint, with key comparisons extended across scales and families.
A single consumer-grade GPU running a vision-language model (VLM) is used to supply missing guidance on pixel-level segmentation, improving segmentation while producing structured, auditable evidence that drives the result and can be inspected on its own.
Hierarchical Block Quantization (HBQ) uses large blocks to maximize efficiency and introduces low-overhead significand (SIG) scaling for second-level quantization and compensates for large-block errors more effectively than prior PoT and INT schemes.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.
Zero-knowledge proofs can verify computing results without rerunning workloads or exposing sensitive data. Magnetar helps bring them into the mainstream with support for standard programming languages, major speed gains, and quantum-resistant security. The post Proof-based general-purpose computing with Magnetar appeared first on Microsoft Research.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
We’re giving developers, enterprises, scientists, and consumers access to Gemini 4 Argon, our latest frontier model.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
