semantic-scholar

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Semantic Scholar Lookup

Semantic Scholar 检索工具

Fast, targeted lookups of paper metadata, citations, and authors using the Semantic Scholar API via

asta papers

commands.

通过

asta papers

命令调用Semantic Scholar API，实现论文元数据、引用信息及作者信息的快速定向检索。

When to Use This Skill

适用场景

User asks for details about a specific paper (by title, DOI, arXiv ID, etc.)
User wants to see papers citing a given work
User asks about an author's papers
User wants a quick keyword search (not a comprehensive report)
User wants to find specific claims, methods, or evidence within paper full text (use
```
snippet-search
```
)
Task requires targeted paper metadata or citation graphs

Not for comprehensive reports - use the Literature Report Generation skill for that.

用户查询特定论文的详情（通过标题、DOI、arXiv ID等）
用户想要查看引用某篇论文的文献
用户查询某作者的论文
用户需要快速关键词搜索（而非综合报告）
用户想要在论文全文中查找特定论点、方法或证据（使用
```
snippet-search
```
）
任务需要定向获取论文元数据或引用图谱

不适用于生成综合报告 - 此类需求请使用文献报告生成Skill。

Related: Find Literature skill

Installation

安装

This skill requires the

asta

CLI:

bash

undefined

本Skill需要

asta

CLI工具：

bash

undefined

Install/reinstall at the correct version

PLUGIN_VERSION=0.17.1 if [ "$(asta --version 2>/dev/null | grep -oE '[0-9]+.[0-9]+.[0-9]+')" != "$PLUGIN_VERSION" ]; then uv tool install --force git+https://github.com/allenai/asta-plugins.git@v$PLUGIN_VERSION fi


**Prerequisites:** Python 3.11+ and [uv package manager](https://docs.astral.sh/uv/)


**前置要求**：Python 3.11+ 和 [uv包管理器](https://docs.astral.sh/uv/)

Available Commands

可用命令

Get Paper Details

获取论文详情

asta papers get - Get metadata for a single paper by ID

bash

asta papers get ARXIV:2005.14165

asta papers get "DOI:10.18653/v1/N18-3011" --fields title,year,authors,abstract

asta papers get CorpusId:215416146 --format text

Supported ID formats:

```
ARXIV:2106.15928
```
```
DOI:10.18653/v1/N18-3011
```
```
CorpusId:215416146
```
```
PMID:19872477
```
```
URL:https://arxiv.org/abs/2106.15928
```

Common fields:

title,abstract,authors,year,venue,citationCount,publicationDate,url,isOpenAccess,fieldsOfStudy

asta papers get - 通过ID获取单篇论文的元数据

bash

asta papers get ARXIV:2005.14165

asta papers get "DOI:10.18653/v1/N18-3011" --fields title,year,authors,abstract

asta papers get CorpusId:215416146 --format text

支持的ID格式：

```
ARXIV:2106.15928
```
```
DOI:10.18653/v1/N18-3011
```
```
CorpusId:215416146
```
```
PMID:19872477
```
```
URL:https://arxiv.org/abs/2106.15928
```

常用字段：

title,abstract,authors,year,venue,citationCount,publicationDate,url,isOpenAccess,fieldsOfStudy

Search Papers

论文搜索

asta papers search - Keyword-based paper search

bash

asta papers search "transformers attention mechanism"

asta papers search "RLHF" --date 2023- --limit 10

asta papers search "neural networks" --fields title,year,abstract,authors

asta papers search "LLM safety" --date 2024-01-01:2024-12-31

Options:

```
--fields
```
: Comma-separated fields to return
```
--limit
```
: Number of results (default 20, max 100)
```
--date
```
: Publication date or year filter. Accepts years (
```
2020
```
,
```
2020-2024
```
,
```
2020-
```
) or date ranges (
```
2024-01-01:2024-12-31
```
). Maps to the S2
```
publicationDateOrYear
```
parameter.
```
--format
```
: Output as
```
json
```
or
```
text
```

asta papers search - 基于关键词的论文搜索

bash

asta papers search "transformers attention mechanism"

asta papers search "RLHF" --date 2023- --limit 10

asta papers search "neural networks" --fields title,year,abstract,authors

asta papers search "LLM safety" --date 2024-01-01:2024-12-31

选项：

```
--fields
```
：指定返回的字段，用逗号分隔
```
--limit
```
：结果数量（默认20，最大100）
```
--date
```
：发表日期或年份筛选。支持年份（
```
2020
```
、
```
2020-2024
```
、
```
2020-
```
）或日期范围（
```
2024-01-01:2024-12-31
```
）。对应S2的
```
publicationDateOrYear
```
参数。
```
--format
```
：输出格式为
```
json
```
或
```
text
```

Snippet Search

片段检索

asta papers snippet-search - Search over paper full text (title, abstract, and body) via the S2 snippet/search API. Returns matching text excerpts alongside paper metadata.

bash

asta papers snippet-search "in-context learning emerges at scale"

asta papers snippet-search "RLHF reward hacking" --date 2023- --limit 10

asta papers snippet-search "sparse mixture of experts" --fields snippet.text,snippet.snippetKind,snippet.section

asta papers snippet-search - 通过S2片段/搜索API检索论文全文（标题、摘要、正文）。返回匹配的文本片段及论文元数据。

bash

asta papers snippet-search "in-context learning emerges at scale"

asta papers snippet-search "RLHF reward hacking" --date 2023- --limit 10

asta papers snippet-search "sparse mixture of experts" --fields snippet.text,snippet.snippetKind,snippet.section

Pin results to papers indexed before a date (useful for reproducible benchmarks)

将结果限定在指定日期前收录的论文（适用于可复现基准测试）

asta papers snippet-search "chain-of-thought" --inserted-before 2024-01-01


The `--fields` option accepts **snippet fields**:

- `snippet.text` - The matched text excerpt (~500 words)
- `snippet.snippetKind` - Source type (e.g., title, abstract, body)
- `snippet.section` - Paper section the snippet came from
- `snippet.snippetOffset` - Character position data (`start`, `end`)
- `snippet.annotations` - Markup including reference mentions and sentence boundaries

If `--fields` is omitted, the default is `snippet.text,snippet.snippetKind`. Paper metadata (corpusId, title, authors, openAccessInfo) and relevance score are always returned regardless of `--fields`.

Options:
- `--fields`: Comma-separated snippet fields to return
- `--date`: Date/year filter, same as standard search
- `--limit`: Max results (default 20, max 1000 — higher ceiling than standard search)
- `--inserted-before`: Only include papers indexed before this date (`YYYY-MM-DD`, `YYYY-MM`, or `YYYY`). Typically used for consistency in benchmarking — pinning a cutoff date ensures the same set of papers is returned across repeated runs, even as new papers are continuously indexed.
- `--format`: Output as `json` or `text`

asta papers snippet-search "chain-of-thought" --inserted-before 2024-01-01


`--fields`选项支持**片段字段**：

- `snippet.text`：匹配的文本片段（约500词）
- `snippet.snippetKind`：来源类型（如标题、摘要、正文）
- `snippet.section`：片段所在的论文章节
- `snippet.snippetOffset`：字符位置数据（`start`、`end`）
- `snippet.annotations`：包含引用提及和句子边界的标记信息

如果省略`--fields`，默认返回`snippet.text,snippet.snippetKind`。无论是否指定`--fields`，论文元数据（corpusId、标题、作者、开放获取信息）和相关性分数都会被返回。

选项：
- `--fields`：指定返回的片段字段，用逗号分隔
- `--date`：日期/年份筛选，与标准搜索一致
- `--limit`：最大结果数（默认20，最大1000——比标准搜索上限更高）
- `--inserted-before`：仅包含指定日期前收录的论文（格式为`YYYY-MM-DD`、`YYYY-MM`或`YYYY`）。通常用于基准测试的一致性——设定截止日期可确保多次运行返回相同的论文集，即使不断有新论文被收录。
- `--format`：输出格式为`json`或`text`

Get Citations

获取引用文献

asta papers citations - Papers that cite a given work

bash

asta papers citations ARXIV:2005.14165

asta papers citations CorpusId:218487638 --limit 20 --format text

Options:

```
--fields
```
: Fields for citing papers
```
--limit
```
: Max results (default 50, max 1000)
```
--format
```
: Output as
```
json
```
or
```
text
```

asta papers citations - 获取引用指定论文的文献

bash

asta papers citations ARXIV:2005.14165

asta papers citations CorpusId:218487638 --limit 20 --format text

选项：

```
--fields
```
：指定引用论文的字段
```
--limit
```
：最大结果数（默认50，最大1000）
```
--format
```
：输出格式为
```
json
```
或
```
text
```

Author Search and Papers

作者搜索与论文查询

asta papers author search - Find authors by name

bash

asta papers author search "Yoav Goldberg"

asta papers author search "Hinton" --limit 5 --format text

asta papers author papers - Get papers by an author

bash

undefined

asta papers author search - 通过姓名查找作者

bash

asta papers author search "Yoav Goldberg"

asta papers author search "Hinton" --limit 5 --format text

asta papers author papers - 获取指定作者的论文

bash

undefined

First, get author ID from search

第一步：通过搜索获取作者ID

asta papers author search "Yoav Goldberg"

Then get their papers using the author ID

第二步：使用作者ID获取其论文

asta papers author papers 1741101 --limit 50

asta papers author papers 1741101 --fields title,year,venue,citationCount


Options:
- `--fields`: Fields to return for papers
- `--limit`: Max results (default 50, max 1000)
- `--format`: Output as `json` or `text`

asta papers author papers 1741101 --limit 50

asta papers author papers 1741101 --fields title,year,venue,citationCount


选项：
- `--fields`：指定返回的论文字段
- `--limit`：最大结果数（默认50，最大1000）
- `--format`：输出格式为`json`或`text`

Output Formats

输出格式

All commands support two output formats:

JSON format (default):

Machine-readable
Complete data structure
Pipe to
```
jq
```
for filtering
Best for programmatic use

Text format (

--format text

Human-readable
Formatted output
Best for quick browsing
Use when showing results to user

所有命令支持两种输出格式：

JSON格式（默认）：

机器可读
完整数据结构
可通过管道传递给
```
jq
```
进行过滤
最适合程序化使用

文本格式（

--format text

）：

人类可读
格式化输出
最适合快速浏览
向用户展示结果时使用

Usage Tips

使用技巧

Efficient Field Selection

高效字段选择

Only request fields you need for faster responses:

bash

undefined

仅请求所需字段以加快响应速度：

bash

undefined

Good - minimal fields for quick browse

Less efficient - many fields slow down response

效率较低 - 请求过多字段会减慢响应速度

asta papers search "deep learning" --fields title,abstract,authors,year,venue,citations,references

undefined

asta papers search "deep learning" --fields title,abstract,authors,year,venue,citations,references

undefined

Date Filtering

日期筛选

Restrict to recent papers when appropriate:

bash

asta papers search "RLHF" --date 2023-2024  # 2023-2024
asta papers search "RLHF" --date 2023-      # 2023 onwards
asta papers search "RLHF" --date -2020      # Before 2020
asta papers search "RLHF" --date 2024-06-01:2024-12-31  # Specific date range

适当时限制为近期论文：

bash

asta papers search "RLHF" --date 2023-2024  # 2023-2024年
asta papers search "RLHF" --date 2023-      # 2023年及以后
asta papers search "RLHF" --date -2020      # 2020年以前
asta papers search "RLHF" --date 2024-06-01:2024-12-31  # 特定日期范围

Piping to jq

管道传递给jq

For complex JSON processing:

bash

undefined

用于复杂JSON处理：

bash

undefined

Extract just titles

仅提取标题

asta papers search "transformers" | jq '.data[].title'

Filter by citation count

按引用数过滤

asta papers search "neural networks" | jq '.data[] | select(.citationCount > 100)'

Get author names

获取作者姓名

asta papers get ARXIV:2005.14165 | jq '.authors[].name'

undefined

asta papers get ARXIV:2005.14165 | jq '.authors[].name'

undefined

Multi-Step Workflows

多步骤工作流

Chain commands for complex queries:

Example 1: Find highly-cited recent papers by an author

bash

undefined

串联命令完成复杂查询：

示例1：查找某作者近期高引用论文

bash

undefined

1. Find author

1. 查找作者

asta papers author search "Geoffrey Hinton" --format text

2. Get their recent papers

2. 获取其近期论文

asta papers author papers 1751273 --fields title,year,citationCount --limit 50 |
jq '.data[].paper | select(.year >= 2020) | select(.citationCount > 100)'


**Example 2: Explore citation network**
```bash

asta papers author papers 1751273 --fields title,year,citationCount --limit 50 |
jq '.data[].paper | select(.year >= 2020) | select(.citationCount > 100)'


**示例2：探索引用网络**
```bash

1. Get paper details

1. 获取论文详情

asta papers get ARXIV:2005.14165

2. Get who cited it

2. 获取引用该论文的文献

asta papers citations ARXIV:2005.14165 --limit 20 --format text

3. Get details on specific citing papers

3. 获取特定引用论文的详情

asta papers get CorpusId:123456789

undefined

asta papers get CorpusId:123456789

undefined

Example Workflows

示例工作流

"Get details for arXiv:2005.14165"

“获取arXiv:2005.14165的详情”

bash

asta papers get ARXIV:2005.14165 --format text

Present the output to user in a readable format.

bash

asta papers get ARXIV:2005.14165 --format text

以可读格式向用户展示输出结果。

"What papers cite the GPT-3 paper?"

“哪些论文引用了GPT-3论文？”

bash

undefined

bash

undefined

GPT-3 paper

GPT-3论文

asta papers citations ARXIV:2005.14165 --limit 50 --format text


Show recent/highly-cited papers from the results.

asta papers citations ARXIV:2005.14165 --limit 50 --format text


从结果中展示近期/高引用的论文。

"Recent papers on RLHF"

“RLHF相关近期论文”

bash

asta papers search "RLHF reinforcement learning from human feedback" \
  --date 2023- \
  --limit 20 \
  --fields title,abstract,year,authors,venue,citationCount \
  --format text

bash

asta papers search "RLHF reinforcement learning from human feedback" \
  --date 2023- \
  --limit 20 \
  --fields title,abstract,year,authors,venue,citationCount \
  --format text

"Papers by Yoav Goldberg"

“Yoav Goldberg的论文”

bash

undefined

bash

undefined

Step 1: Find author

步骤1：查找作者

asta papers author search "Yoav Goldberg" --format text

Step 2: Get their papers (using author ID from step 1)

步骤2：使用步骤1得到的作者ID获取其论文

asta papers author papers 1741101
--fields title,year,venue,citationCount
--limit 50
--format text

undefined

asta papers author papers 1741101
--fields title,year,venue,citationCount
--limit 50
--format text

undefined

"Find evidence of 'chain-of-thought' reasoning"

“查找'思维链'推理的相关证据”

bash

undefined

bash

undefined

Snippet search finds mentions in paper bodies, not just titles/abstracts

片段检索可在论文正文中查找提及内容，而非仅标题/摘要

asta papers snippet-search "chain-of-thought reasoning improves performance"
--limit 15

Or use standard search for paper-level results

或使用标准搜索获取论文级结果

asta papers search "chain-of-thought reasoning"
--fields title,abstract,year,authors
--limit 15

Then examine specific papers with 'asta papers get'

然后使用'asta papers get'查看特定论文详情

undefined

undefined

Response Presentation

结果展示

When showing results to users:

For single paper:

**Title** (Year)
Authors: [author list]
Venue: [venue name]
Citations: [count]

[Abstract]

URL: [Semantic Scholar link]

For paper lists:

Found [N] papers:

1. **Paper Title** - Author et al. (Year) - [Venue] - [X citations]
2. **Another Paper** - ...
...

For snippet results (

snippet-search

Found [N] snippet results:

1. **Paper Title** - Author et al.
   Score: 0.95
   Snippet (abstract): "...matching text excerpt..."
2. **Another Paper** - ...
...

For citations:

Found [N] papers citing this work:

Recent citations:
1. [Paper 1] (2024) - [citations]
2. [Paper 2] (2023) - [citations]
...

向用户展示结果时：

单篇论文：

**标题**（年份）
作者：[作者列表]
发表平台：[平台名称]
引用数：[数量]

[摘要]

链接：[Semantic Scholar链接]

论文列表：

找到[N]篇论文：

1. **论文标题** - 作者等（年份） - [发表平台] - [X次引用]
2. **另一篇论文** - ...
...

片段检索结果（

snippet-search

）：

找到[N]条片段结果：

1. **论文标题** - 作者等
   分数：0.95
   片段（摘要）："...匹配的文本片段..."
2. **另一篇论文** - ...
...

引用文献：

找到[N]篇引用该论文的文献：

近期引用：
1. [论文1]（2024） - [引用数]
2. [论文2]（2023） - [引用数]
...

Best Practices

最佳实践

Use
```
--format text
```
when showing results directly to user
Use JSON output when you need to process or filter results
Start with small limits, increase if needed
Only fetch fields you'll actually use
Use
```
snippet-search
```
when searching for specific claims, methods, or evidence within paper bodies
Use
```
search
```
for topic-level paper discovery
For comprehensive research, suggest Literature Report Generation skill instead
Provide Semantic Scholar URLs when helpful (
```
https://semanticscholar.org/paper/{paperId}
```
)

直接向用户展示结果时使用
```
--format text
```
需要处理或过滤结果时使用JSON输出
先使用较小的结果数量限制，必要时再增加
仅获取实际需要的字段
查找论文正文中的特定论点、方法或证据时使用
```
snippet-search
```
主题级论文发现使用
```
search
```
如需全面研究，建议使用文献报告生成Skill

必要时提供Semantic Scholar链接（

https://semanticscholar.org/paper/{paperId}

）

API Key

API密钥

The commands use the

ASTA_TOOL_KEY

environment variable if available. Most queries work without a key, but a key provides higher rate limits.

命令会使用环境变量

ASTA_TOOL_KEY

（如果已设置）。大多数查询无需密钥即可使用，但密钥可提供更高的请求速率限制。

semantic-scholar

Original

Translation

Semantic Scholar Lookup

Semantic Scholar 检索工具

When to Use This Skill

适用场景

Related: Find Literature skill

相关工具：Find Literature Skill

Installation

安装

Install/reinstall at the correct version

Install/reinstall at the correct version

Available Commands

可用命令

Get Paper Details

获取论文详情

Search Papers

论文搜索

Snippet Search

片段检索

Pin results to papers indexed before a date (useful for reproducible benchmarks)

将结果限定在指定日期前收录的论文（适用于可复现基准测试）

Get Citations

获取引用文献

Author Search and Papers

作者搜索与论文查询

First, get author ID from search

第一步：通过搜索获取作者ID

Then get their papers using the author ID

第二步：使用作者ID获取其论文

Output Formats

输出格式

Usage Tips

使用技巧

Efficient Field Selection

高效字段选择

Good - minimal fields for quick browse

推荐 - 仅请求快速浏览所需的少量字段

Less efficient - many fields slow down response

效率较低 - 请求过多字段会减慢响应速度

Date Filtering

日期筛选

Piping to jq

管道传递给jq

Extract just titles

仅提取标题

Filter by citation count

按引用数过滤

Get author names

获取作者姓名

Multi-Step Workflows

多步骤工作流

1. Find author

1. 查找作者

2. Get their recent papers

2. 获取其近期论文

1. Get paper details

1. 获取论文详情

2. Get who cited it

2. 获取引用该论文的文献

3. Get details on specific citing papers

3. 获取特定引用论文的详情

Example Workflows

示例工作流

"Get details for arXiv:2005.14165"

“获取arXiv:2005.14165的详情”

"What papers cite the GPT-3 paper?"

“哪些论文引用了GPT-3论文？”

GPT-3 paper

GPT-3论文

"Recent papers on RLHF"

“RLHF相关近期论文”

"Papers by Yoav Goldberg"

“Yoav Goldberg的论文”

Step 1: Find author

步骤1：查找作者

Step 2: Get their papers (using author ID from step 1)

步骤2：使用步骤1得到的作者ID获取其论文

"Find evidence of 'chain-of-thought' reasoning"