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

    Neuronpedia

    Jacobian LensNEW
    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. 2-RES-JB
    11. 14309
    Prev
    Next
    INDEX
    Explanations

    phrases related to political campaigns

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "riding"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ilities
    -0.81
    ormon
    -0.74
    mson
    -0.71
    mercial
    -0.70
    chie
    -0.69
    argon
    -0.67
    rencies
    -0.67
    vernment
    -0.67
    erent
    -0.66
    elman
    -0.66
    POSITIVE LOGITS
    bike
    0.99
     riding
    0.94
     bicycles
    0.90
     rode
    0.89
     ride
    0.88
     rides
    0.88
     bikes
    0.86
     motorcycles
    0.86
     Riding
    0.85
     ridden
    0.81
    Activations Density 0.014%

    No Known Activations