© 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. 63840
    Prev
    Next
    INDEX
    Explanations

    terms related to risk, harm, and adverse effects

    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
    ÅĻÃŃzenÃŃ
    -0.14
    LAS
    -0.13
    âĢ¦.↵↵
    -0.13
    estre
    -0.13
    خت
    -0.12
     Kab
    -0.12
    [âĢ¦
    -0.12
    isation
    -0.12
    liness
    -0.12
    andas
    -0.12
    POSITIVE LOGITS
    ated
    0.15
    ating
    0.14
    리ì¦Ī
    0.14
    alled
    0.13
    MMdd
    0.13
    inating
    0.13
    inated
    0.13
    -toggler
    0.13
    lients
    0.12
    é¡
    0.12
    Activations Density 0.255%

    No Known Activations