프롬프트 구조화, 장문맥 Attention, 양방향 Diffusion, 로컬 AI 인프라

9 techniques to generate better LLM outputs: An instruction alone, with no examples, schema, or stated constraints, is the default way to generate an LLM response. This is zero-shot prompting, and it serves as the baseline. But when outputs aren't good enough, the first thing you should try is to improve the prompt, not a model upgrade. There are 8 prompting techniques that you can use: - Few-shot prompting: Show the model a few input-output examples. It picks up the pattern and applies it to new inputs. - Chain-of-thought (CoT): Ask for step-by-step reasoning. Breaks complex problems into smaller, verifiable steps. - Prompt hierarchy: Use system, developer, and user prompts at different authority levels. Higher levels override lower ones. - Role-specific prompting: Assign a persona like "You are a security researcher." The model conditions on a different slice of its training data. - Negative prompting: Specify what the LLM should NOT do. "Do not use marketing language" works as a hard constraint. - JSON prompting: Define a JSON schema in the prompt. The LLM returns structured output matching that schema. - Attentive reasoning queries (ARQ): Replace free-form CoT with structured, domain-specific questions. In testing, it hit 90.2% instruction adherence vs. 81.5% for direct prompting. - Verbalized sampling: Explicitly prompt the LLM to generate multiple responses with probability scores. Recovers the output diversity that RLHF alignment suppresses. They stack well together. Few-shot + CoT. JSON + Negative prompting. ARQ is essentially a structured version of CoT for agentic conversations. Of course, better outputs will also come from better context, tool access, and retrieval. But these 9 techniques live entirely in the prompt and involve working with the same model, infrastructure, and setup. You are just structuring the prompt differently to get better outputs from an LLM. Once the prompt is doing its job, the next lever is how much compute the model spends before it answers. I covered that separately, walking through chain of thought, best-of-N, self-refinement, and tree search, and why each of them only keeps paying off when you have a reliable way to verify the output. Read it below. PS: This visual was inspired by ByteByteGo's post on a similar topic. I added some latest research-driven techniques like ARQ, Verbalized sampling, and JSON prompting.

Sliding-window beats linear attention "In this work, we show that Sliding Window Attention (SWA) with sinks performs as well or better than post-trained Linear Attention models. We observe this across multiple LLMs on various downstream tasks. For long-context reasoning tasks (Needle-in-a-Haystack and BABILong), SWA achieves massively higher performance (2 to 10 times higher than linear attention). SWA requires no post-training, is extremely fast, and requires low memory; therefore, making it an extremely cheap and reliable solution. To reduce inference memory cost, we strongly recommend switching to SWA instead of post-training linear models. Linear attention models may have shown some promise, but they likely require to be trained from scratch or extensive post-training in order to even match SWA." paper link: https:// arxiv.org/abs/2608.28444

There and Back Again: Bidirectional Diffusion Bridges for Multimodality Translation "We propose BIT: Bidirectional Image-Text Diffusion Bridges. In contrast to previous approaches, BIT starts directly from text and interpolates into images, providing (1) a source-aware generative path that enables diverse and flexible sampling algorithms; and (2) an endpoint-conditioned process that can be traversed from image to text, providing a unified, bidirectional generative framework." "BIT represents a paradigm shift by challenging the assumption that T2I generation needs to start from a Gaussian noise distribution." project page: https:// bit-diffusion.github.io code: https:// github.com/gabeguo/bit_di ffusion … paper link: https:// arxiv.org/abs/2608.27885

The Mac mini may be becoming something very different from a desktop computer: A dedicated local AI node. For coding agents, private document workflows, and persistent local assistants, unified-memory Macs occupy an interesting gap between laptops and cloud GPUs. Local AI is becoming infrastructure.

A commercial director's take on Wan3.0. Nick Tasker @SLAIstudio walks through the 30-second ad he made with it — the concept, the craft, and why prompt adherence matters to a director more than anything else. 0:08 Intro 1:37 Where the idea came from 5:07 Workflow 10:11 The advantages of Wan 3 in AI ads 12:44 Tips for creators Watch ↓

Everyone can generate a SPIDER-MAN with Hy4 preview

-World 1-1, two prompts

Alibaba Cloud has officially launched its first cloud region in Brazil, featuring two new data centers in São Paulo. A major expansion in Latin America!

This milestone reinforces our commitment to supporting Brazilian businesses with robust, resilient, and secure local cloud infrastructure. Learn more: https:// alibabacloud.com/en/press-room/ alibaba-cloud-launch-first-cloud-region-in-brazil … #AlibabaAI