© 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. 9530
    Prev
    Next
    INDEX
    Explanations

    words related to someone being implicated or participating in various activities

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "involved."

    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
    eb
    -0.68
    itone
    -0.66
    stand
    -0.62
     testament
    -0.62
     Preferences
    -0.62
    wards
    -0.62
    helm
    -0.59
    lock
    -0.59
    ipes
    -0.59
    velt
    -0.58
    POSITIVE LOGITS
     involved
    3.69
    involved
    2.44
     implicated
    1.85
     involve
    1.58
     concerned
    1.57
     engaged
    1.49
     embroiled
    1.48
     involvement
    1.40
     involves
    1.35
     responsible
    1.33
    Activations Density 0.024%

    No Known Activations