File size: 2,718 Bytes
fe4a4f7
 
 
ec448a7
fe4a4f7
 
 
ec448a7
182832e
fe4a4f7
374da48
fe4a4f7
 
 
fec9632
 
 
fe4a4f7
 
 
374da48
fec9632
374da48
 
 
fec9632
 
 
 
 
 
 
 
 
 
 
5eea7c0
 
 
ec448a7
 
374da48
5eea7c0
 
374da48
 
 
 
 
 
fe4a4f7
374da48
fe4a4f7
 
 
374da48
fec9632
374da48
fe4a4f7
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
import streamlit as st

st.set_page_config(
    page_title="Pathfinder",
    page_icon="πŸ‘‹",
)

# st.write("# Welcome to Pathfinder! πŸ‘‹")
st.image('local_files/pathfinder_logo.png',caption="Pathfinder: LLM enabled literature search")
st.sidebar.success("Select a function above.")
st.sidebar.markdown("Current functions include visualizing papers in the arxiv embedding, searching for similar papers to an input paper or prompt phrase, or answering quick questions.")

st.markdown(
    """
    Pathfinder (formerly called arXiv+GPT) is a framework for searching and
    visualizing papers on the [arXiv](https://arxiv.org/) using the context
    sensitivity from modern large language models (LLMs) to better link paper contexts.

    **πŸ‘ˆ Select a tool from the sidebar** to see some examples
    of what this framework can do!

    ### Tool summary:
    - `Paper search` looks for relevant papers given an arxiv id or a question.
    - `Arxiv embedding` shows the landscape of current galaxy evolution papers (astro-ph.GA)
    - `QA sources` brings it all together to give concise answers to questions with primary sources and relevant papers.

    This is not meant to be a replacement to existing tools like the
    [ADS](https://ui.adsabs.harvard.edu/),
    [arxivsorter](https://www.arxivsorter.org/), but rather a supplement to find papers
    that otherwise might be missed during a literature survey.
    It is also only trained on astro-ph.GA (astrophysics of galaxies) papers currently,
    if you are interested in extending it please reach out!

    The image below shows a representation of all the astro-ph.GA papers that can be explored in more detail
    using the `Arxiv embedding` page. The papers tend to cluster together by similarity, and result in an
    atlas that shows well studied (forests) and currently uncharted areas (water).
    """
)

# st.image('https://drive.google.com/uc?id=1yQQCdlgnFzi-_yOMplGIqEyPKJhIsZpO&export=download')
st.image('local_files/galaxy_worldmap_kiyer-min.png')

st.markdown(
    """
    ### Coming soon:
    - [AstroLLaMA](https://huggingface.co/spaces/universeTBD/astrollama) embeddings!
    - export results
    - daily updates to repo
    - other fields apart from `astro-ph.GA`

    ### Want to learn more?
    - Check out `AstroLLaMA` [paper](https://huggingface.co/papers/2309.06126)
    - Check out `chaotic_neural` [(link)](http://chaotic-neural.readthedocs.io/)
    - Jump into our [documentation](https://docs.streamlit.io)
    - Contribute!

    Pathfinder is developed and maintained by [UniverseTBD](https://universetbd.org/). Updates on [huggingface](https://huggingface.co/universeTBD) or [twitter](https://twitter.com/universe_tbd).

"""
)