docs: link report on arXiv and fix resource icons

README.md

@@ -7,7 +7,7 @@ tags:
 
 # Index-Translate-9B
 
-[Online demo](https://index-translate.bilibili.com/) · [GitHub](https://github.com/bilibili/Index-Translate) · [Technical report](https://github.com/bilibili/Index-Translate/blob/main/docs/Index_Translate_Series_Technical_Report.pdf) · [Hugging Face collection](https://huggingface.co/collections/IndexTeam/index-translate) · [ModelScope collection](https://www.modelscope.cn/collections/IndexTeam/Index-Translate)
+<p><a style="display:inline" href="https://index-translate.bilibili.com/">🌐 Online demo</a> · <a style="display:inline" href="https://github.com/bilibili/Index-Translate"><img src="https://raw.githubusercontent.com/bilibili/Index-Translate/main/docs/assets/github.svg" width="16" height="16" alt="" style="display:inline;vertical-align:middle;margin:0"> GitHub</a> · <a style="display:inline" href="https://arxiv.org/abs/2609.40181">📚 Technical report</a> · <a style="display:inline" href="https://huggingface.co/collections/IndexTeam/index-translate">🤗 Hugging Face collection</a> · <a style="display:inline" href="https://www.modelscope.cn/collections/IndexTeam/Index-Translate"><img src="https://raw.githubusercontent.com/bilibili/Index-Translate/main/docs/assets/modelscope.svg" width="16" height="16" alt="" style="display:inline;vertical-align:middle;margin:0"> ModelScope collection</a></p>
 
 **Index-Translate-9B** is the 9B-parameter multilingual text model of the **Index-Translate** family, built on Qwen3.5. It supports text translation across **150 languages** and follows translation instructions involving terminology, formatting, style, structure, context, and output length.
 
@@ -107,7 +107,7 @@ These answer-accuracy results (%) use matched evaluation subsets. Translation tr
 - Text translation evaluation uses **greedy decoding and a 16,384-token context window**, except WMT26, which uses zero-shot Chinese instructions, a **3,584-token context**, and a **1,536-token output limit**. WMT26 reference-free GPT-5.6-Sol scores are averaged over 20 target languages.
 - IFMTBench starts with 7,344 examples. Preprocessing removes 280 single-constraint examples with reference or language-label issues, leaving **7,064 inputs**; **7,063 receive valid judgments**. See the [preprocessing note](https://github.com/bilibili/Index-Translate/blob/main/docs/evaluation.md#ifmtbench-preprocessing).
 
-The **1,024-token output budget in the inference client below is a usage default**, not the output budget used to produce every reported evaluation result. Detailed protocols and valid-judgment counts are in the [technical report](https://github.com/bilibili/Index-Translate/blob/main/docs/Index_Translate_Series_Technical_Report.pdf). The report describes instTrans and MEME; benchmark release plans are tracked in the [GitHub TODO](https://github.com/bilibili/Index-Translate#todo).
+The **1,024-token output budget in the inference client below is a usage default**, not the output budget used to produce every reported evaluation result. Detailed protocols and valid-judgment counts are in the [technical report](https://arxiv.org/abs/2609.40181). The report describes instTrans and MEME; benchmark release plans are tracked in the [GitHub TODO](https://github.com/bilibili/Index-Translate#todo).
 
 ## Inference
 
@@ -256,7 +256,7 @@ Scores vary across languages and tasks. Combined constraints and low-resource di
   institution={Index LLM Team},
   year={2026},
   month={September},
-  url={https://github.com/bilibili/Index-Translate/blob/main/docs/Index_Translate_Series_Technical_Report.pdf}
+  url={https://arxiv.org/abs/2609.40181}
 }
 ```