© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 28121
    Prev
    Next
    INDEX
    Explanations

    elements related to criticism and commentary on societal issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     novamente
    -0.17
    ropoda
    -0.15
     partially
    -0.15
    ambre
    -0.14
    eyen
    -0.14
    HeaderCode
    -0.14
    è«
    -0.14
    ency
    -0.14
    rsa
    -0.14
    çĶ³åįļ
    -0.14
    POSITIVE LOGITS
     Ãł
    0.16
     ("
    0.16
     involving
    0.15
     (“
    0.15
    ä
    0.15
    oga
    0.15
    chez
    0.15
     ÑĤипа
    0.14
    íķĻ기
    0.14
     Featuring
    0.14
    Activations Density 0.216%

    No Known Activations