© 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. 0-RES-JB
    11. 13476
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    plate
    -0.85
    orgetown
    -0.85
    plates
    -0.83
    uckland
    -0.78
    itect
    -0.77
    ãĥ¼ãĥĨãĤ£
    -0.75
    arta
    -0.73
    mson
    -0.70
    atable
    -0.69
    abetic
    -0.69
    POSITIVE LOGITS
     Creed
    0.65
    ophy
    0.65
     manure
    0.63
     Bore
    0.63
     Sector
    0.62
     recommendation
    0.62
    rity
    0.62
     Bermuda
    0.60
    exe
    0.59
     viruses
    0.59
    Activations Density 0.000%

    No Known Activations

    This feature has no known activations.