© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 11-RES-JB
    11. 21935
    Prev
    Next
    INDEX
    Explanations

    mentions or citations of articles, blogs, or social media posts authored by specific individuals

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "via" indicating references or sources

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    adal
    -0.87
    inary
    -0.76
    Enlarge
    -0.75
    stal
    -0.73
    lesh
    -0.72
    thal
    -0.69
    é¾įå
    -0.67
    aed
    -0.64
    $,
    -0.63
    bane
    -0.62
    POSITIVE LOGITS
     Wikimedia
    1.06
     email
    0.99
     Skype
    0.96
     Shutterstock
    0.91
     Twitter
    0.85
     Email
    0.84
     youtube
    0.82
     YouTube
    0.81
     Flickr
    0.80
     Gmail
    0.78
    Activations Density 0.043%

    No Known Activations