Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 1-RES-JB
    6. 19846
    Prev
    Next
    INDEX
    Explanations

    references to the media platform "BuzzFeed"

    oai_token-act-pair · gpt-3.5-turbo

    references to the "Buzzfeed" media outlet

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    esthetic
    -0.81
     discipl
    -0.81
    xual
    -0.75
     dedication
    -0.73
    izations
    -0.72
    semble
    -0.71
     restitution
    -0.70
     nonviolent
    -0.70
    egal
    -0.70
     compassionate
    -0.68
    POSITIVE LOGITS
     Buzz
    1.23
    feed
    1.17
    Buzz
    0.96
    ards
    0.95
    Feed
    0.87
    arded
    0.86
    arro
    0.86
    oola
    0.83
     buzz
    0.82
    vine
    0.80
    Activations Density 0.003%

    No Known Activations