© 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
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 22736
    Prev
    Next
    INDEX
    Explanations

    phrases related to saving lives and preventing harm

    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
    lop
    -0.14
     shrink
    -0.13
    ajor
    -0.13
    ilde
    -0.13
    ually
    -0.12
    ائÙĬ
    -0.12
     rendez
    -0.12
    aise
    -0.12
    342
    -0.12
    thur
    -0.12
    POSITIVE LOGITS
     lives
    1.00
     Lives
    0.82
     life
    0.77
    life
    0.63
     Life
    0.59
     Leben
    0.58
    Life
    0.56
    -life
    0.56
     жизни
    0.55
     LIFE
    0.55
    Activations Density 0.048%

    No Known Activations