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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 3-RES-JB
    11. 11437
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Redditor
    -0.87
    ãĥĨãĤ£
    -0.79
    ij士
    -0.74
    ĸ
    -0.70
    hemer
    -0.69
     Valiant
    -0.68
    æĸ¹
    -0.67
    é¾įåĸļ士
    -0.65
    Mist
    -0.63
    Cla
    -0.62
    POSITIVE LOGITS
    iston
    0.68
    angs
    0.66
     teenagers
    0.65
     girls
    0.64
    alls
    0.63
    eret
    0.62
     outlets
    0.62
    ultz
    0.62
     rapists
    0.61
    enge
    0.61
    Activations Density 0.000%

    No Known Activations

    This feature has no known activations.