© 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. 43274
    Prev
    Next
    INDEX
    Explanations

    references to family structures and issues related to familial or societal naming conventions

    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
    ts
    -0.18
     is
    -0.15
     Was
    -0.14
     was
    -0.14
    Was
    -0.14
    oid
    -0.14
    ted
    -0.14
    ournals
    -0.14
    swith
    -0.13
    ,is
    -0.13
    POSITIVE LOGITS
     are
    0.27
     aren
    0.23
     such
    0.20
     ARE
    0.20
    ifter
    0.18
     were
    0.18
    Are
    0.17
    _are
    0.17
     Are
    0.17
     weren
    0.17
    Activations Density 0.395%

    No Known Activations