© 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 EvalsExportsinterp-engineNEWAPI 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. 20859
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "ow" followed by a single digit number, possibly indicating emotion or sensation

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "ow" and its variations

    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
    ually
    -0.79
    $$$$
    -0.66
    uring
    -0.65
     capsule
    -0.65
     suspic
    -0.62
    ãĥ¼ãĥĨ
    -0.60
    uration
    -0.60
    uate
    -0.59
    URA
    -0.59
     Mahm
    -0.59
    POSITIVE LOGITS
    orld
    1.69
    itsch
    1.27
    orks
    1.20
    riter
    1.17
    olves
    1.12
    idth
    1.12
    iak
    1.09
    rites
    1.07
    atts
    1.05
    ards
    1.03
    Activations Density 0.049%

    No Known Activations