---
id: "/tags/primary-source/"
title: "Primary Source"
creator: null
creator_url: null
creator_x: null
url: "https://skillissue.dev/tags/primary-source/"
level: null
level_max: null
category: null
subcategory: null
region: null
license: null
license_scope: null
access: null
models: null
source_date: null
date_added: null
---

# Primary Source

[Beginner](/levels/beginner/index.md) [Intermediate](/levels/intermediate/index.md) [Advanced](/levels/advanced/index.md) [Expert](/levels/expert/index.md)

8 resources · page 1 of 1

- 

### [Text-to-text prompt guide (Alibaba Cloud)](/prompting/prompting-fundamentals/alibaba-cloud-text-to-text-prompt-guide/index.md)

by [Alibaba Cloud](https://www.alibabacloud.com/help/en/model-studio/prompt-engineering-guide)

Learn to specify context, goals and output format using the model publisher’s own prompting guidance.

Beginner · CN · closed resource · en 

- 

### [Indirect prompt injection in the real world: how people manipulate neural networks](/security-and-red-teaming/prompt-injection/kaspersky-indirect-prompt-injection-field-study/index.md)

by [Vladislav Tushkanov (Kaspersky)](https://securelist.com/indirect-prompt-injection-in-the-wild/113295/)

Read a Russia-origin security team’s field study of indirect prompt injection and the implications for defensive risk assessment.

Beginner · RU · closed resource · en 

- 

### [Yandex Research: natural language processing](/frontier/research-feeds/yandex-research-nlp-feed/index.md)

by [Yandex Research](https://research.yandex.com/research-areas/natural-language-processing)

Follow primary Yandex research on language models, inference, distributed training and related datasets.

Intermediate · RU · closed resource · en 

- 

### [JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models](/security-and-red-teaming/jailbreak-research/chao-jailbreakbench/index.md)

by [Patrick Chao, Edoardo Debenedetti, Alexander Robey, Maksym Andriushchenko and colleagues](https://arxiv.org/abs/2404.01318v5)

Learn why jailbreak studies need consistent threat models, scoring and reproducible comparisons.

Advanced · West · closed resource · en 

- 

### [GigaChat Family: Efficient Russian Language Modeling Through Mixture of Experts Architecture](/model-training/pretraining/gigachat-family-moe-training-report/index.md)

by [GigaChat team / SaluteDevices](https://arxiv.org/abs/2506.09440v1)

Study a Russian-origin model team’s account of mixture-of-experts architecture, pretraining and evaluation.

Advanced · RU · open resource · en 

- 

### [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](/security-and-red-teaming/jailbreak-research/mazeika-harmbench/index.md)

by [Mantas Mazeika, Long Phan and colleagues](https://arxiv.org/abs/2402.04249v2)

Study a standardized approach to red-team evaluation and robust refusal without copying attack recipes.

Advanced · West · open resource · en 

- 

### [Qwen3 Technical Report](/frontier/model-releases/qwen3-technical-report/index.md)

by [Qwen Team / Alibaba](https://arxiv.org/abs/2505.09388v1)

Read the model team’s account of Qwen3’s dense and MoE releases and unified thinking and non-thinking modes.

Advanced · CN · closed resource · en 

- 

### [Qwen3Guard Technical Report](/security-and-red-teaming/safety-and-alignment-evals/qwen3guard-technical-report/index.md)

by [Haiquan Zhao and colleagues / Qwen Team](https://github.com/QwenLM/Qwen3Guard)

Understand Qwen’s multilingual safety classifiers and the distinction between whole-response and streaming moderation.

Advanced · CN · closed resource · en 

[Plain-text Markdown copy](/tags/primary-source/index.md)