LIVE · 05/12
agent-fw★RELlangchain-core==1.4.0langchain-core==1.4.0[langchain-releases]vscodeRELVisual Studio Code 1.120 Release NotesVisual Studio Code 1.120 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.119 Release NotesVisual Studio Code 1.119 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.118 Release NotesVisual Studio Code 1.118 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.117 Release NotesVisual Studio Code 1.117 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.116 Release NotesVisual Studio Code 1.116 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.115 Release NotesVisual Studio Code 1.115 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.114 Release NotesVisual Studio Code 1.114 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.113 Release NotesVisual Studio Code 1.113 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.112 Release NotesVisual Studio Code 1.112 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.111 Release NotesVisual Studio Code 1.111 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.110 Release NotesVisual Studio Code 1.110 Release Notes[vscode-updates]vscodeRELVisual Studio Code 1.109 Release NotesVisual Studio Code 1.109 Release Notes[vscode-updates]cursorRELCursor in Microsoft TeamsCursor in Microsoft Teams[cursor-changelog]researchFlashSVD v1.5: Making Low-Rank Transformers Inference Actually FastFlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast[arxiv-cs-ai]researchSingle-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack SuccessSingle-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success[arxiv-cs-ai]researchProactBench: Beyond What The User Asked ForProactBench: Beyond What The User Asked For[arxiv-cs-ai]researchRethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient?Rethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient?[arxiv-cs-ai]researchWhere Reliability Lives in Vision-Language Models: A Mechanistic Study of Attention, Hidden States, and Causal CircuitsWhere Reliability Lives in Vision-Language Models: A Mechanistic Study of Attention, Hidden States, and Causal Circuits[arxiv-cs-ai]researchSpatial Priming Outperforms Semantic Prompting: A Grid-Based Approach to Improving LLM Accuracy on Chart Data ExtractionSpatial Priming Outperforms Semantic Prompting: A Grid-Based Approach to Improving LLM Accuracy on Chart Data Extraction[arxiv-cs-ai]researchAuto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative CriteriaAuto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria[arxiv-cs-ai]researchEmbeddings for Preferences, Not SemanticsEmbeddings for Preferences, Not Semantics[arxiv-cs-ai]researchOn Distinguishing Capability Elicitation from Capability Creation in Post-Training: A Free-Energy PerspectiveOn Distinguishing Capability Elicitation from Capability Creation in Post-Training: A Free-Energy Perspective[arxiv-cs-ai]researchMemQ: Integrating Q-Learning into Self-Evolving Memory Agents over Provenance DAGsMemQ: Integrating Q-Learning into Self-Evolving Memory Agents over Provenance DAGs[arxiv-cs-ai]
Today 56
Total 519
Major 23
Active sources 25/51
Updated just now
langchain-core==1.4.0
Featured · 注目 Featured

langchain-core==1.4.0 langchain-core==1.4.0

Changes since langchain-core==0.3.86 chore(infra): merge v1.4 into master ( #37350 ) chore: bump urllib3 from 2.6.3 to 2.7.0 in /libs/core ( #37329 ) fix(core): avoid eager pydantic.v1 import in @depr

release langchain-releases · 9h ago
Daily Summary

今日の更新

Today's Updates

Today 56 ▼ 81%
Yesterday 288
7-day 502
Last 7 days
34
48
26
14
36
288
56
05/06 05/07 05/08 05/09 05/10 05/11 05/12
Last 7 days article counts
DateCount
2026-05-0634
2026-05-0748
2026-05-0826
2026-05-0914
2026-05-1036
2026-05-11288
2026-05-1256
Top categories · 7d
主要な更新 Top stories 05/12 · 10 件
  1. 01 agent-fw ★ REL langchain-core==1.4.0 langchain-core==1.4.0 Changes since langchain-core==0.3.86 chore(infra): merge v1.4 into master ( #37350 ) chore: bump urllib3 from 2.6.3 to 2.7.0 in /libs/core ( #37329 ) fix(core): avoid eager pydantic.v1 import in @depr [langchain-releases]
  2. 02 vscode REL +11 Visual Studio Code 1.120 Release Notes Visual Studio Code 1.120 Release Notes Monthly release notes for VS Code 1.120. [vscode-updates]
  3. 03 cursor REL Cursor in Microsoft Teams Cursor in Microsoft Teams Cursor is now available in Microsoft Teams. [cursor-changelog]
  4. 04 research FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast arXiv:2605.08314v1 Announce Type: cross Abstract: SVD-based Low-rank compression reduces transformer parameters and nominal FLOPs, but these savings often translate poorly into real LLM serving speedu [arxiv-cs-ai]
  5. 05 research Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success arXiv:2605.09070v1 Announce Type: cross Abstract: Many jailbreak attack research papers report attack success rates for a limited number of parameter settings, even though there are many combinations [arxiv-cs-ai]
  6. 06 claude Claude Opus 4.7 Claude Opus 4.7 [anthropic-news]
  7. 07 claude Claude Design Anthropic Labs Claude Design Anthropic Labs [anthropic-news]
  8. 08 local-llm DDLだけでは不十分？LLMで「実態」に即したER図を再構築するための検証プロセス (no English title) はじめに ドキュメントが存在しない（あるいは陳腐化している）レガシーシステムのDB設計を復元したいとき、LLMは強力な味方です。DDLを貼り付けて「ER図を作って」と頼めば、数分で整ったドキュメントが出てきます。 ところが、ある業務システム [qiita-llm]
  9. 09 claude Javaの複雑な処理フローを1発で抽出する！設計書生成プロンプトに必ず入れるべき4つの指示 (no English title) はじめに 既存のJavaコードから設計書を起こしたくて、AIに「このメソッドの仕様をまとめて」と投げてみたことがあります。帰ってくる出力は正確ではあるのですが、なんとなく物足りない。具体的なコード値が出てこない、例外スローはあるのにメッセー [qiita-claude]
  10. 10 local-llm Claude Codeのスレッド名を日本語にしたくて内部構造を調べたら、過去スレッドが全部消えた話【解決策求む】 (no English title) Claude Code（デスクトップアプリ）を使っていると、サイドバーのスレッド名がすべて英語で生成されます😇 「GAS spreadsheet date navigation」とか「Fix SSH permission denied e [qiita-llm]
🔥 Today's Top 3 importance × recency
  1. langchain-core==1.4.0 langchain-core==1.4.0 langchain-releases 9h ago
  2. CodeQL 2.25.3 adds Swift 6.3 support CodeQL 2.25.3 adds Swift 6.3 support github-changelog 3d ago
  3. Visual Studio Code 1.120 Release Notes Visual Studio Code 1.120 Release Notes vscode-updates just now

Timeline 519 total · page 1/18

TODAY 30 entries
NEW paper research 1m ago · arxiv-cs-ai

FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast

EN arXiv:2605.08314v1 Announce Type: cross Abstract: SVD-based Low-rank compression reduces transformer parameters and nominal FLOPs, but these savings often translate poorly into real LLM serving speedu

EN arXiv:2605.08314v1 Announce Type: cross Abstract: SVD-based Low-rank compression reduces transformer parameters and nominal FLOPs, but these savings often translate poorly into real LLM serving speedu

#arxiv#paper
arxiv.org →
FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast og
NEW paper research 1m ago · arxiv-cs-ai

Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success

EN arXiv:2605.09070v1 Announce Type: cross Abstract: Many jailbreak attack research papers report attack success rates for a limited number of parameter settings, even though there are many combinations

EN arXiv:2605.09070v1 Announce Type: cross Abstract: Many jailbreak attack research papers report attack success rates for a limited number of parameter settings, even though there are many combinations

#arxiv#paper
arxiv.org →
Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success og
NEW paper research 1m ago · arxiv-cs-ai

ProactBench: Beyond What The User Asked For ProactBench: Beyond What The User Asked For

EN arXiv:2605.09228v1 Announce Type: cross Abstract: Most LLM benchmarks score how well a model responds to explicit requests. They leave unmeasured a different conversational ability: noticing and actin

EN arXiv:2605.09228v1 Announce Type: cross Abstract: Most LLM benchmarks score how well a model responds to explicit requests. They leave unmeasured a different conversational ability: noticing and actin

#arxiv#paper
arxiv.org →
ProactBench: Beyond What The User Asked For og
NEW paper research 1m ago · arxiv-cs-ai

Rethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient? Rethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient?

EN arXiv:2605.10848v1 Announce Type: cross Abstract: Does a lexical retriever suffice as large language models (LLMs) become more capable in an agentic loop? This question naturally arises when building

EN arXiv:2605.10848v1 Announce Type: cross Abstract: Does a lexical retriever suffice as large language models (LLMs) become more capable in an agentic loop? This question naturally arises when building

#agent#arxiv#paper
arxiv.org →
Rethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient? og
NEW paper research 1m ago · arxiv-cs-ai

Where Reliability Lives in Vision-Language Models: A Mechanistic Study of Attention, Hidden States, and Causal Circuits Where Reliability Lives in Vision-Language Models: A Mechanistic Study of Attention, Hidden States, and Causal Circuits

EN arXiv:2605.08200v1 Announce Type: new Abstract: A pervasive intuition holds that vision-language models (VLMs) are most trustworthy when their attention maps look sharp: concentrated attention on the

EN arXiv:2605.08200v1 Announce Type: new Abstract: A pervasive intuition holds that vision-language models (VLMs) are most trustworthy when their attention maps look sharp: concentrated attention on the

#arxiv#paper
arxiv.org →
NEW paper research 1m ago · arxiv-cs-ai

Spatial Priming Outperforms Semantic Prompting: A Grid-Based Approach to Improving LLM Accuracy on Chart Data Extraction Spatial Priming Outperforms Semantic Prompting: A Grid-Based Approach to Improving LLM Accuracy on Chart Data Extraction

EN arXiv:2605.08220v1 Announce Type: new Abstract: The automated extraction of data from scientific charts is a critical task for large-scale literature analysis. While multimodal Large Language Models (

EN arXiv:2605.08220v1 Announce Type: new Abstract: The automated extraction of data from scientific charts is a critical task for large-scale literature analysis. While multimodal Large Language Models (

#arxiv#paper
arxiv.org →
NEW paper research 1m ago · arxiv-cs-ai

Auto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria Auto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria

EN arXiv:2605.08354v1 Announce Type: new Abstract: Aligning multimodal generative models with human preferences demands reward signals that respect the compositional, multi-dimensional structure of human

EN arXiv:2605.08354v1 Announce Type: new Abstract: Aligning multimodal generative models with human preferences demands reward signals that respect the compositional, multi-dimensional structure of human

#arxiv#paper
arxiv.org →
healthy harness v0.1 powered by Astro + Claude Opus Cloudflare Pages