发布日期:2026-05-25
收录条目:10
1. Build a Complete Langfuse Observability and Evaluation Pipeline for Tracing, Prompt Management, Scoring, and Experiments
- 来源:MarkTechPost
- 发布时间:2026-05-24 23:03 UTC
- 链接:https://www.marktechpost.com/2026/05/24/build-a-complete-langfuse-observability-and-evaluation-pipeline-for-tracing-prompt-management-scoring-and-experiments/
摘要:In this tutorial, we implement the Langfuse (an open-source LLM engineering platform) pipeline for tracing, prompt management, scoring, datasets, and experiments. We build a complete workflow that works with either a rea
2. StepFun Releases StepAudio 2.5 Realtime: An End-to-End Voice Model with Roleplay-Specific RLHF and Paralinguistic Comprehension
- 来源:MarkTechPost
- 发布时间:2026-05-24 22:51 UTC
- 链接:https://www.marktechpost.com/2026/05/24/stepfun-releases-stepaudio-2-5-realtime-an-end-to-end-voice-model-with-roleplay-specific-rlhf-and-paralinguistic-comprehension/
摘要:StepFun, the Shanghai-based AI lab, released StepAudio 2.5 Realtime in May 2026 — an end-to-end real-time speech large language model with fully customizable persona capabilities. The model connects via a WebSocket API,
3. Hackers are learning to exploit chatbot ‘personalities’
- 来源:The Verge AI
- 发布时间:2026-05-24 12:00 UTC
- 链接:https://www.theverge.com/column/935545/hackers-ai-chatbots
摘要:This is The Stepback, a weekly newsletter breaking down one essential story from the tech world. For more on AI mischief, follow Robert Hart. The Stepback arrives in our subscribers' inboxes at 8AM ET. Opt in for The Ste
4. Microsoft Research Releases Webwright: A Terminal-Native Web Agent Framework That Scores 60.1% on Odysseys, Up from Base GPT-5.4’s 33.5%
- 来源:MarkTechPost
- 发布时间:2026-05-24 08:56 UTC
- 链接:https://www.marktechpost.com/2026/05/24/microsoft-research-releases-webwright-a-terminal-native-web-agent-framework-that-scores-60-1-on-odysseys-up-from-base-gpt-5-4s-33-5/
摘要:Microsoft Research introduces Webwright, a terminal-native browser agent framework that replaces click-trace web automation with reusable Playwright scripts. Using a single agent loop across three modules and roughly 1,0
5. NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule
- 来源:MarkTechPost
- 发布时间:2026-05-24 07:42 UTC
- 链接:https://www.marktechpost.com/2026/05/24/nvidia-ai-releases-gated-deltanet-2-a-linear-attention-layer-that-decouples-erase-and-write-in-the-delta-rule/
摘要:Linear attention squeezes the unbounded KV cache into a fixed-size recurrent state, but editing that memory without scrambling existing associations is hard. Prior delta-rule models like Gated DeltaNet and KDA use one sc
6. Tencent Open-Sources TencentDB Agent Memory: A 4-Tier Local Memory Pipeline for AI Agents
- 来源:MarkTechPost
- 发布时间:2026-05-23 19:31 UTC
- 链接:https://www.marktechpost.com/2026/05/23/tencent-open-sources-tencentdb-agent-memory-a-4-tier-local-memory-pipeline-for-ai-agents/
摘要:Tencent has open-sourced TencentDB Agent Memory, a fully local memory system for AI agents released under the MIT license. The project pairs symbolic short-term memory, which offloads verbose tool logs into a compact Mer
7. Build a SuperClaude Framework Workflow with Commands, Agents, Modes, and Session Memory
- 来源:MarkTechPost
- 发布时间:2026-05-23 19:05 UTC
- 链接:https://www.marktechpost.com/2026/05/23/build-a-superclaude-framework-workflow-with-commands-agents-modes-and-session-memory/
摘要:In this tutorial, we build an advanced workflow using the SuperClaude Framework as a structured layer on top of the Anthropic API. The post Build a SuperClaude Framework Workflow with Commands, Agents, Modes, and Session
8. Google’s new anything-to-anything AI model is wild
- 来源:The Verge AI
- 发布时间:2026-05-23 11:00 UTC
- 链接:https://www.theverge.com/tech/936507/gemini-omni-hands-on-deepfake-ai-video
摘要:Last year I deepfaked my kid's stuffed animal to make it look like his plush deer was on vacation. It was an experiment to see if I could re-create the events depicted in a Gemini ad Google was running, and I never showe
9. Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification
- 来源:MarkTechPost
- 发布时间:2026-05-23 10:32 UTC
- 链接:https://www.marktechpost.com/2026/05/23/nous-research-releases-contrastive-neuron-attribution-cna-sparse-mlp-circuit-steering-without-sae-training-or-weight-modification/
摘要:Nous Research releases Contrastive Neuron Attribution (CNA), a method that identifies and ablates sparse MLP neuron circuits to steer LLM behavior — no sparse autoencoder training, no weight modification, and no degradat
10. Perplexity Open-Sources Bumblebee: A Read-Only Supply-Chain Scanner for Developer Endpoints
- 来源:MarkTechPost
- 发布时间:2026-05-23 08:17 UTC
- 链接:https://www.marktechpost.com/2026/05/23/perplexity-open-sources-bumblebee-a-read-only-supply-chain-scanner-for-developer-endpoints/
摘要:Perplexity has open-sourced Bumblebee, an internal security tool it uses to protect the developer systems behind its search product, Comet, and Computer. Bumblebee is a read-only inventory collector for macOS and Linux d