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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 30038
    Prev
    Next
    INDEX
    Explanations

    instances of reported abuse or neglect in institutional settings

    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
    _Debug
    -0.14
    ả
    -0.13
    ÑģилÑĮ
    -0.13
     assassin
    -0.13
    ossil
    -0.12
    ém
    -0.12
    ViewById
    -0.12
    าศ
    -0.12
    XC
    -0.12
     Äijá»ķ
    -0.12
    POSITIVE LOGITS
     cases
    0.66
     case
    0.58
    cases
    0.50
     Cases
    0.50
     instances
    0.46
     incidents
    0.46
    case
    0.43
     casos
    0.43
    Cases
    0.42
     Case
    0.42
    Activations Density 0.133%

    No Known Activations