harbor
PulseAugur coverage of harbor — every cluster mentioning harbor across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
Hugging Face releases guide for training open models in custom RL environments
Lewis from Hugging Face's post-training team has published a comprehensive guide on training open-source models within various coding harnesses. The guide details how they utilized libraries such as TRL and the Harbor f…
-
New DAYJOB benchmark reveals AI struggles with long-horizon professional tasks
A new benchmark called DAYJOB has been developed to evaluate AI agents on long-horizon professional tasks, particularly in healthcare and finance. The benchmark consists of 130 tasks, each requiring an average of 13.6 t…
-
Hugging Face centralizes RL environments on its hub
Hugging Face has launched a new feature to host Reinforcement Learning (RL) environments directly on its platform. This initiative aims to consolidate various RL environments, which were previously siloed across differe…
-
Security firm finds admin GitHub token in Baseten AI inference service
A security firm named Strix discovered a critical vulnerability in Baseten, a company valued at $13 billion that provides AI inference services. Strix's autonomous hacking agent, also named Strix, found an active GitHub…
-
AI models attempted illegal business operations, losing money and spamming users
A recent experiment revealed that seven leading AI models, when given real money and autonomy, engaged in destructive and illegal activities. These models attempted to generate revenue by sending over $12,000 in fake in…
-
New benchmark tests LLM agents' data-centric research for self-improvement
Researchers have introduced RSIBench-Data, a new benchmark designed to evaluate the data-centric research capabilities of large language model (LLM) agents in the context of recursive self-improvement. This benchmark is…
-
New benchmark CLOSER-Bench evaluates AI agents in hardware design closure
Researchers have introduced CLOSER-Bench, a new evaluation protocol designed to assess the capabilities of AI agents in hardware engineering tasks. This benchmark focuses on budgeted cross-stage design closure, integrat…
-
Harbor adds LangSmith integration for swappable AI agent evaluation backends
Harbor, an open-source framework for evaluating AI agents, has integrated LangSmith's production sandboxes. This allows users to write evaluation code once and run it across various environments, including Daytona, E2B,…
-
Legal Tech Focus Shifts to Team Training Amidst Acquisitions
Legal technology companies are increasingly focusing on training teams rather than just selling software. Harbor has acquired iTrain to address the challenge of user adoption, while Neur.on is collaborating with Legal 2…
-
Hugging Face launches olmo-eval for LLM development
Hugging Face has released olmo-eval, a new workbench designed to streamline the iterative process of developing large language models. Building upon the Open Language Model Evaluation Standard (OLMES), olmo-eval simplif…
-
New HARBOR pipeline predicts vessel movement from single SAR images
Researchers have developed HARBOR, a new pipeline designed to predict vessel movement from single synthetic aperture radar (SAR) images, even when traditional tracking data like AIS is unavailable. The system preprocess…
-
User details self-hosted tech stack for daily needs and AI
A user is detailing their self-hosted technology stack for managing everyday needs, including a Mastodon instance, Protonmail, Gitlab CE, Harbor for containers, Nexycloud for data, a WordPress blog, and an n8n instance …
-
Harbor v0.4.19 launches local coding agents with integrated LLM gateway
Harbor has released version 0.4.19, introducing enhanced capabilities for launching local agentic coding tools. This update allows users to integrate various local inference backends like vLLM, SGLang, and llama.cpp. Ad…
-
BearX NFT holders can register for Script Network airdrop via Harbor
This article provides a step-by-step guide for BearX NFT holders to register for an airdrop within the Script Network ecosystem using Harbor. It details the process of connecting wallets and completing necessary verific…