© 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. 8-RES-JB
    11. 10482
    Prev
    Next
    INDEX
    Explanations

    the title "Legend of Korra"

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of "Legend of" with varying suffixes, indicating it focuses on the "Legend of Korra" series specifically

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    abus
    -0.74
    earcher
    -0.71
    tan
    -0.70
    ters
    -0.63
     Parenthood
    -0.61
    chens
    -0.61
    ktop
    -0.60
    sels
    -0.60
    pex
    -0.59
    chy
    -0.57
    POSITIVE LOGITS
    arily
    1.31
    naire
    1.02
    arium
    0.97
    aries
    0.87
    ocl
    0.84
    ous
    0.84
    naires
    0.79
     Gord
    0.78
     tales
    0.76
    Legend
    0.76
    Activations Density 0.030%

    No Known Activations