© 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. 0-RES-JB
    11. 5207
    Prev
    Next
    INDEX
    Explanations

    phrases related to clicking on links or buttons for specific actions or purposes

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "FOR."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ess
    -0.60
    esses
    -0.60
    lav
    -0.57
     joints
    -0.57
    bum
    -0.55
    rys
    -0.55
     incidents
    -0.54
    lev
    -0.53
    metal
    -0.52
    ese
    -0.51
    POSITIVE LOGITS
     FOR
    2.97
    FOR
    2.36
     TO
    1.53
     OF
    1.48
     ABOUT
    1.47
     IN
    1.45
     WITH
    1.44
     FROM
    1.44
     WHEN
    1.42
     BY
    1.36
    Activations Density 0.017%

    No Known Activations