© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 63790
    Prev
    Next
    INDEX
    Explanations

    titles and roles related to authority figures in legal and medical contexts

    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
    jvu
    -0.16
    NSNotification
    -0.15
    ¶ļ
    -0.15
    [--
    -0.14
    zcze
    -0.14
    AVIS
    -0.14
    кÑģп
    -0.14
    antz
    -0.14
    .glide
    -0.14
    sheets
    -0.14
    POSITIVE LOGITS
    Dr
    0.15
    plat
    0.15
    athom
    0.15
     Dr
    0.15
    ONO
    0.14
    cona
    0.14
     designate
    0.14
    isin
    0.14
    ysters
    0.14
    orate
    0.14
    Activations Density 0.027%

    No Known Activations