© 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 EvalsExportsinterp-engineNEWAPI 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. 22331
    Prev
    Next
    INDEX
    Explanations

    references to antisemitism and its relationship to other forms of racism

    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
    letal
    -0.16
    hevik
    -0.15
    üy
    -0.14
    .gwt
    -0.14
    ساÙĨ
    -0.14
    iminal
    -0.14
    agnar
    -0.14
     Fou
    -0.14
    çłĤ
    -0.14
    PYTHON
    -0.14
    POSITIVE LOGITS
     anti
    0.60
     Anti
    0.56
     antis
    0.55
    antis
    0.54
    anti
    0.54
    Anti
    0.51
    ANTI
    0.41
     Ant
    0.37
     Jew
    0.37
    Ant
    0.31
    Activations Density 0.080%

    No Known Activations