From efdbf44ce8f43d1c548fd0c7190e75bcec47b229 Mon Sep 17 00:00:00 2001 From: bits-bytes-nn Date: Tue, 17 Nov 2026 20:51:44 +0900 Subject: [PATCH 1/2] chore: record the URLs of six published posts The five posts merged in #129-#133 and the prompt-injection survey from 2026-11-02, so validate-site.sh fails if any of them stops being built. --- script/published-urls.txt | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/script/published-urls.txt b/script/published-urls.txt index e7c74cb..2b652a9 100644 --- a/script/published-urls.txt +++ b/script/published-urls.txt @@ -40,8 +40,11 @@ /paper-reviews/finetuning/2022/12/15/constitutional-ai-harmlessness-from-ai-feedback.html /paper-reviews/finetuning/2023/05/23/qlora-efficient-finetuning-of-quantized-llms.html /paper-reviews/finetuning/2023/05/29/direct-preference-optimization-your-language-model-is-secretly-a-reward-model.html +/paper-reviews/finetuning/2024/06/17/optimizing-instructions-and-demonstrations-for-multi-stage-language-model.html +/paper-reviews/finetuning/2025/07/25/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.html /paper-reviews/language-models/2023/02/27/llama-open-and-efficient-foundation-language-models.html /paper-reviews/language-models/2023/07/18/llama-2-open-foundation-and-fine-tuned-chat-models.html +/paper-reviews/language-models/2023/10/05/dspy-compiling-declarative-language-model-calls-into-self-improving-pipelines.html /paper-reviews/language-models/2023/10/10/mistral-7b.html /paper-reviews/language-models/2024/01/05/deepseek-llm-scaling-open-source-language-models-with-longtermism.html /paper-reviews/language-models/2024/01/08/mixtral-of-experts.html @@ -65,5 +68,8 @@ /paper-reviews/retrieval-augmented-generation/2024/04/24/from-local-to-global-a-graph-rag-approach-to-query-focused-summarization.html /paper-reviews/retrieval-augmented-generation/2024/10/08/lightrag-simple-and-fast-retrieval-augmented-generation.html /paper-reviews/retrieval-augmented-generation/2025/01/20/zep-a-temporal-knowledge-graph-architecture-for-agent-memory.html +/paper-summaries/agentic-ai/2023/05/23/rewoo-decoupling-reasoning-from-observations-for-efficient-augmented-language-summary.html +/paper-summaries/agentic-ai/2024/02/01/executable-code-actions-elicit-better-llm-agents-summary.html /paper-summaries/agentic-ai/2025/11/26/toolorchestra-elevating-intelligence-via-efficient-model-and-tool-orchestration-summary.html +/paper-surveys/agentic-ai/2026/11/02/securing-llm-agents-from-prompt-injection-to-mcp-and-defenses.html /tech-guides/agentic-ai/2026/10/23/strands-agents-sdk-the-agent-loop-hooks-and-production-controls.html From e64dcd06d4ff316c12109b094bbe29ce05aeae19 Mon Sep 17 00:00:00 2001 From: bits-bytes-nn Date: Tue, 17 Nov 2026 20:52:44 +0900 Subject: [PATCH 2/2] fix(post): move the MIPROv2 and GEPA reviews to Language-Models They are prompt optimizers, like the DSPy review there, not finetuning. The topic's cover follows, and each old Finetuning URL, live for an hour, redirects to the new one (recorded, so the stub must keep building). --- ...ompt-evolution-can-outperform-reinforcement-learning.md | 7 +++++-- ...ns-and-demonstrations-for-multi-stage-language-model.md | 7 +++++-- script/published-urls.txt | 2 ++ 3 files changed, 12 insertions(+), 4 deletions(-) diff --git a/_posts/2026-11-17-gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.md b/_posts/2026-11-17-gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.md index c567b28..b6b5823 100644 --- a/_posts/2026-11-17-gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.md +++ b/_posts/2026-11-17-gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.md @@ -7,10 +7,13 @@ date: 2025-07-25 17:42:32 paper_author: "UC Berkeley" paper_doi: "10.48550/arXiv.2507.19457" description: "GEPA는 실행 트레이스를 자연어로 성찰해 지시문을 고치고 파레토 프런티어로 후보를 고르는 프롬프트 옵티마이저입니다. Qwen3 8B 여섯 과제에서 최대 35배 적은 롤아웃으로 GRPO를 평균 6% 앞서며, 구조와 실험 결과를 정리합니다." -categories: ["Paper Reviews", "Finetuning"] +categories: ["Paper Reviews", "Language-Models"] tags: ["Reflective-Prompt-Evolution", "Pareto-Based-Candidate-Selection", "Natural-Language-Reflection", "Prompt-Optimization", "Sample-Efficient-Optimization", "Reasoning-Models"] -cover: /assets/images/finetuning.jpg +cover: /assets/images/language-models.jpg use_math: true +# First published under the Finetuning topic. +redirect_from: + - /paper-reviews/finetuning/2025/07/25/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.html --- ## TL;DR diff --git a/_posts/2026-11-17-optimizing-instructions-and-demonstrations-for-multi-stage-language-model.md b/_posts/2026-11-17-optimizing-instructions-and-demonstrations-for-multi-stage-language-model.md index d1839d7..82cffc8 100644 --- a/_posts/2026-11-17-optimizing-instructions-and-demonstrations-for-multi-stage-language-model.md +++ b/_posts/2026-11-17-optimizing-instructions-and-demonstrations-for-multi-stage-language-model.md @@ -7,10 +7,13 @@ date: 2024-06-17 16:12:03 paper_author: "Stanford University" paper_doi: "10.48550/arXiv.2406.11695" description: "MIPRO는 다단계 LM 프로그램에서 모듈별 지시문과 시연 예시를 중간 레이블 없이 함께 최적화하는 DSPy 옵티마이저이며, Llama-3-8B를 쓴 일곱 과제 중 다섯에서 기준 옵티마이저를 앞섰습니다. 제안 전략, 크레딧 할당 방식, 실험 결과를 정리합니다." -categories: ["Paper Reviews", "Finetuning"] +categories: ["Paper Reviews", "Language-Models"] tags: ["MIPRO-Optimizer", "Multi-Stage-LM-Programs", "Prompt-Optimization", "Bayesian-Surrogate-Model", "Instruction-Proposal", "Agentic-AI"] -cover: /assets/images/finetuning.jpg +cover: /assets/images/language-models.jpg use_math: true +# First published under the Finetuning topic. +redirect_from: + - /paper-reviews/finetuning/2024/06/17/optimizing-instructions-and-demonstrations-for-multi-stage-language-model.html --- ## TL;DR diff --git a/script/published-urls.txt b/script/published-urls.txt index 2b652a9..5fd0428 100644 --- a/script/published-urls.txt +++ b/script/published-urls.txt @@ -52,12 +52,14 @@ /paper-reviews/language-models/2024/01/25/deepseek-coder-when-the-large-language-model-meets-programming-the-rise-of-code-intelligence.html /paper-reviews/language-models/2024/02/05/deepseekmath-pushing-the-limits-of-mathematical-reasoning-in-open-language-models.html /paper-reviews/language-models/2024/05/07/deepseek-v2-a-strong-economical-and-efficient-mixture-of-experts-language-model.html +/paper-reviews/language-models/2024/06/17/optimizing-instructions-and-demonstrations-for-multi-stage-language-model.html /paper-reviews/language-models/2024/07/31/gemma-2-improving-open-language-models-at-a-practical-size.html /paper-reviews/language-models/2024/07/31/the-llama-3-herd-of-models.html /paper-reviews/language-models/2024/11/22/tulu-3-pushing-frontiers-in-open-language-model-post-training.html /paper-reviews/language-models/2024/12/27/deepseek-v3-technical-report.html /paper-reviews/language-models/2025/01/22/deepseek-r1-incentivizing-reasoning-capability-in-llms-via-reinforcement-learning.html /paper-reviews/language-models/2025/05/14/qwen3-technical-report.html +/paper-reviews/language-models/2025/07/25/gepa-reflective-prompt-evolution-can-outperform-reinforcement-learning.html /paper-reviews/language-models/2025/07/28/kimi-k2-open-agentic-intelligence.html /paper-reviews/language-models/2025/12/02/deepseek-v3.2-pushing-the-frontier-of-open-large-language-models.html /paper-reviews/multimodal-learning/2023/12/19/gemini-a-family-of-highly-capable-multimodal-models.html