Skip to main content

Post summary

Svelte

The post summary component summarizes various content and associated meta data into a highly configurable component.

Classes

Section titled Classes
Class Parent Description
.s-post-summary N/A Base parent container for a post summary
.s-post-summary__answered .s-post-summary Adds the styling necessary for a question with accepted answers
.s-post-summary__deleted .s-post-summary Adds the styling necessary for a deleted post
.s-post-summary--sm-hide .s-post-summary Hides the stats container on small screens
.s-post-summary--sm-show .s-post-summary Shows the stats container on small screens
.s-post-summary--answers .s-post-summary Container for the post summary answers
.s-post-summary--answer .s-post-summary--answers Container for the post summary answer
.s-post-summary--answer__accepted .s-post-summary--answer Adds the styling necessary for an accepted answer
.s-post-summary--content .s-post-summary Container for the post summary content
.s-post-summary--content-meta .s-post-summary--content A container for post meta data, things like tags and user cards.
.s-post-summary--content-type .s-post-summary--content Container for the post summary content type
.s-post-summary--excerpt .s-post-summary--content Container for the post summary excerpt
.s-post-summary--stats .s-post-summary Container for the post summary stats
.s-post-summary--stats-answers .s-post-summary--stats Container for the post summary answers
.s-post-summary--stats-bounty .s-post-summary--stats Container for the post summary bounty
.s-post-summary--stats-item .s-post-summary--stats A genericcontainer for views, comments, read time, and other meta data which prepends a separator icon.
.s-post-summary--stats-votes .s-post-summary--stats Container for the post summary votes
.s-post-summary--tags .s-post-summary Container for the post summary tags
.s-post-summary--title .s-post-summary Container for the post summary title
.s-post-summary--title-link .s-post-summary--title Link styling for the post summary title
.s-post-summary--title-icon .s-post-summary--title Icon styling for the post summary title

Examples

Section titled Examples

Base

Section titled Base

Use the post summary component to provide a concise summary of a question, article, or other content.

<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">
        <div class="s-post-summary--stats-votes">…</div>
        <div class="s-post-summary--stats-answers">…</div>
    </div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">
                <div class="s-user-card--group" href="…">
                    <a class="s-avatar" href="…">
                        <img class="s-avatar--image" src="…">
                        <span class="v-visible-sr">…</span>
                    </a>
                    <span class="s-user-card--username">…</span>
                    <ul class="s-user-card--group">
                        <li class="s-user-card--rep">
                            <span class="s-bling s-bling__rep s-bling__sm">
                                <span class="v-visible-sr">reputation bling</span>
                            </span>
                            …
                        </li>
                    </ul>
                    <span>
                        <a class="s-user-card--time" title="…" data-controller="s-tooltip" href="…">
                            <time>…</time>
                        </a>
                    </span>
                </div>
            </div>
            <div class="s-post-summary--stats s-post-summary--sm-show">
                <div class="s-post-summary--stats-votes">
                    {% icon "Vote16Up" %}
                    …
                </div>
                <div class="s-post-summary--stats-answers">
                    {% icon "Answer16" %}
                    …
                    <span class="v-visible-sr">answers</span>
                </div>
            </div>
            <div class="s-post-summary--stats-item">… views</div>
        </div>
        <div class="s-post-summary--title">
            <a class="s-post-summary--title-link" href="…">…</a>
        </div>
        <p class="s-post-summary--excerpt v-truncate3">…</p>
        <div class="s-post-summary--tags">
            <a class="s-tag" href="…">…</a>
            …
        </div>
    </div>
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Answered

Section titled Answered

Add the .s-post-summary__answered modifier class to indicate that the post has an accepted answer. 

<div class="s-post-summary s-post-summary__answered">
    …
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Bountied

Section titled Bountied

Include the .s-post-summary--stats-bounty element to indicate that the post has a bounty. 

<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">
        <div class="s-post-summary--stats-votes">…</div>
        <div class="s-post-summary--stats-answers">…</div>
        <div class="s-post-summary--stats-bounty">
            +50 <span class="v-visible-sr">bounty</span>
        </div>
    </div>
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">
                <div class="s-post-summary--stats-votes">…</div>
                <div class="s-post-summary--stats-answers">…</div>
                <div class="s-post-summary--stats-bounty">
                    +50 <span class="v-visible-sr">bounty</span>
                </div>
            </div>
        </div>
        …
    </div>
    …
</div>
+24 votes
1 answers
+ 50
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
+ 50
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Ignored

Section titled Ignored

Including an ignored tag will automatically apply custom ignored styling to the post summary.

<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-tag s-tag__ignored" href="…">…</a>
            …
        </div>
    </div>
</div>
+24 votes
1 answers
+ 50
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
+ 50
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

Ignored tag
retrieval-augmented-generation langchain llm vector-database ai

Watched

Section titled Watched

Including a watched tag will automatically apply custom watched styling to the post summary.

<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-tag s-tag__watched" href="…">…</a>
            …
        </div>
    </div>
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

Watched tag
retrieval-augmented-generation langchain llm vector-database ai

Deleted

Section titled Deleted

Include the .s-post-summary__deleted modifier class applies custom deleted styling to the post summary.

<div class="s-post-summary s-post-summary__deleted">
    …
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

State badges

Section titled State badges

Include the appropriate state badge to indicate the current state of the post.

<!-- Draft -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--sm-show">
            <span class="s-badge s-badge__sm s-badge__info">
                {% icon "Compose" %} Draft
            </span>
        </div>
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">…</div>
            <div class="s-post-summary--stats-item">… views</div>
            <span class="s-badge s-badge__info ml-auto s-post-summary--sm-hide">
                {% icon "Compose" %} Draft
            </span>
        </div>
        …
    </div>
</div>

<!-- Review -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--sm-show">
            <span class="s-badge s-badge__sm s-badge__warning">
                {% icon "Eye" %} Review
            </span>
        </div>
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">…</div>
            <div class="s-post-summary--stats-item">… views</div>
            <span class="s-badge s-badge__warning ml-auto s-post-summary--sm-hide">
                {% icon "Eye" %} Review
            </span>
        </div>
        …
    </div>
</div>

<!-- Closed -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--sm-show">
            <span class="s-badge s-badge__sm s-badge__danger">
                {% icon "Flag" %} Closed
            </span>
        </div>
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">…</div>
            <div class="s-post-summary--stats-item">… views</div>
            <span class="s-badge s-badge__danger ml-auto s-post-summary--sm-hide">
                {% icon "Flag" %} Closed
            </span>
        </div>
        …
    </div>
</div>

<!-- Archived -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--sm-show">
            <span class="s-badge s-badge__sm">
                {% icon "Document" %} Archived
            </span>
        </div>
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">…</div>
            <div class="s-post-summary--stats-item">… views</div>
            <span class="s-badge ml-auto s-post-summary--sm-hide">
                {% icon "Document" %} Archived
            </span>
        </div>
        …
    </div>
</div>

<!-- Pinned -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--sm-show">
            <span class="s-badge s-badge__sm s-badge__tonal">
                {% icon "Key" %} Pinned
            </span>
        </div>
        <div class="s-post-summary--content-meta">
            <div class="s-user-card s-user-card__sm">…</div>
            <div class="s-post-summary--stats s-post-summary--sm-show">…</div>
            <div class="s-post-summary--stats-item">… views</div>
            <span class="s-badge s-badge__danger ml-auto s-post-summary--sm-hide">
                {% icon "Key" %} Pinned
            </span>
        </div>
        …
    </div>
</div>
+24 votes
1 answers
Draft
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
Draft
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
Review
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
Review
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
Closed
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
Closed
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
Archived
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
Archived
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
Pinned
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
Pinned
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Content types

Section titled Content types

Include the appropriate content type badge to indicate the type of content the post represents.

<!-- Announcement -->
<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-post-summary--content-type" href="#">
                {% icon "Document" %} Announcement
            </a>            …
        </div>
    </div>
</div>

<!-- How-to guide -->
<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-post-summary--content-type" href="#">
                {% icon "Document" %} How to guide
            </a>            …
        </div>
    </div>
</div>

<!-- Knowledge article -->
<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-post-summary--content-type" href="#">
                {% icon "Document" %} Knowledge article
            </a>            …
        </div>
    </div>
</div>

<!-- Policy -->
<div class="s-post-summary">
    …
    <div class="s-post-summary--content">
        …
        <div class="s-post-summary--tags">
            <a class="s-post-summary--content-type" href="#">
                {% icon "Document" %} Policy
            </a>            …
        </div>
    </div>
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

Announcement retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

How to guide retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

Knowledge article retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

Policy retrieval-augmented-generation langchain llm vector-database

Excerpt sizes

Section titled Excerpt sizes

Classes

Section titled Classes

Post summaries can be shown without an excerpt or with an excerpt with one, two, or three lines of text. Exclude the excerpt container to hide the excerpt or apply the appropriate truncation class to the excerpt container. See also Truncation.

Class Description
.v-truncate1 Truncates the excerpt to 1 lines of text.
.v-truncate2 Truncates the excerpt to 2 lines of text.
.v-truncate3 Truncates the excerpt to 3 lines of text.

Examples

Section titled Examples
<!-- No excerpt -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">…</div>
</div>

<!-- Small excerpt -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        …
        <p class="s-post-summary--excerpt v-truncate1">…</p>
        …
    </div>
</div>

<!-- Medium excerpt -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        …
        <p class="s-post-summary--excerpt v-truncate2">…</p>
        …
    </div>
</div>

<!-- Large excerpt -->
<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        …
        <p class="s-post-summary--excerpt v-truncate3">…</p>
        …
    </div>
</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?
retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Small container

Section titled Small container

Post summaries adapt to their container size. When shown with a container smaller than 448px, the post summary renders with a compact layout.

<div class="s-post-summary">…</div>
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
+24 votes
1 answers
+ 50
SofiaAlc SofiaAlc
  • reputation bling 1
24
1 answers
+ 50
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database

Answers

Section titled Answers

Answers to a question can be shown in a post summary. Include the .s-post-summary--answers container to show the answers.

For accepted answers, add the .s-post-summary--answer__accepted modifier class and display the Accepted answer text and icon as shown in the example below.

<div class="s-post-summary">
    <div class="s-post-summary--stats s-post-summary--sm-hide">…</div>
    <div class="s-post-summary--content">
        <div class="s-post-summary--content-meta">…</div>
        <div class="s-post-summary--title">…</div>
        <p class="s-post-summary--excerpt v-truncate3">…</p>
        <div class="s-post-summary--tags">…</div>
        <div class="s-post-summary--answers">
            <div class="s-post-summary--answer s-post-summary--answer__accepted">
                <div class="s-post-summary--content-meta">
                    <div class="s-user-card s-user-card__sm">
                        <div class="s-user-card--group" href="…">
                            <a class="s-avatar" href="…">
                                <img class="s-avatar--image" src="…">
                                <span class="v-visible-sr">…</span>
                            </a>
                            <span class="s-user-card--username">…</span>
                        </div>
                        <ul class="s-user-card--group">
                            <li class="s-user-card--rep">
                                <span class="s-bling s-bling__rep s-bling__sm">
                                    <span class="v-visible-sr">reputation bling</span>
                                </span>
                                …
                            </li>
                        </ul>
                        <span>
                            <a class="s-user-card--time" title="…" data-controller="s-tooltip" href="…">
                                <time>…</time>
                            </a>
                        </span>
                    </div>
                    <div class="s-post-summary--stats">
                        <div class="s-post-summary--stats-votes">
                            {% icon "Vote16Up" %}
                            …
                        </div>
                        <div class="s-post-summary--stats-answers">
                            {% icon "Answer16Fill" %}
                            Accepted answer
                        </div>
                    </div>
                </div>
                <p class="s-post-summary--excerpt">…</p>
            </div>
            <div class="s-post-summary--answer">
                <div class="s-post-summary--content-meta">
                    <div class="s-user-card s-user-card__sm">
                        <div class="s-user-card--group" href="…">
                            <a class="s-avatar" href="…">
                                <img class="s-avatar--image" src="…">
                                <span class="v-visible-sr">…</span>
                            </a>
                            <span class="s-user-card--username">…</span>
                        </div>
                        <ul class="s-user-card--group">
                            <li class="s-user-card--rep">
                                <span class="s-bling s-bling__rep s-bling__sm">
                                    <span class="v-visible-sr">reputation bling</span>
                                </span>
                                …
                            </li>
                        </ul>
                        <span>
                            <a class="s-user-card--time" title="…" data-controller="s-tooltip" href="…">
                                <time>…</time>
                            </a>
                        </span>
                    </div>
                    <div class="s-post-summary--stats">
                        <div class="s-post-summary--stats-votes">
                            {% icon "Vote16Up" %}
                            …
                        </div>
                    </div>
                </div>
                <p class="s-post-summary--excerpt">…</p>
            </div>
        </div>
    </div>
</div>
+24 votes
2 answers
SofiaAlc SofiaAlc
  • reputation bling 1
24
2 answers
98 views
How to reduce hallucinations and improve source relevance in a RAG pipeline?

I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.

retrieval-augmented-generation langchain llm vector-database
SofiaAlc SofiaAlc
  • reputation bling 1
24
Accepted answer
I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.
SofiaAlc SofiaAlc
  • reputation bling 1
24
I have built a Retrieval-Augmented Generation (RAG) system using LangChain, a vector database, and an open-source LLM. While it works reasonably well, the model often hallucinates answers or cites sources that are only tangentially related to the user's query. My chunking strategy is set to a chunk size of 1000 tokens, which seems to be the sweet spot for the model.
How’d we do?
Anonymously upvote, downvote, or send additional feedback below.
Edit File an issue
Deploys by Netlify