© 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. 11-RES-JB
    11. 14541
    Prev
    Next
    INDEX
    Explanations

    references to the TV show "The Legend of Korra."

    oai_token-act-pair · gpt-3.5-turbo

    repeated phrases that express relation or belonging

    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
     acknow
    -0.74
     compliment
    -0.68
     privile
    -0.68
    rouse
    -0.67
    estate
    -0.65
    afety
    -0.65
     plaster
    -0.64
    priority
    -0.64
     closely
    -0.64
     externally
    -0.63
    POSITIVE LOGITS
     Zelda
    1.27
     Korra
    1.25
     Mana
    0.98
     Conan
    0.97
     Ember
    0.96
     Warcraft
    0.95
     Darkness
    0.95
     Tanks
    0.93
     Atlantis
    0.92
     Heroes
    0.91
    Activations Density 0.058%

    No Known Activations