© 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. 14493
    Prev
    Next
    INDEX
    Explanations

    phrases associated with prevention and measures to mitigate negative outcomes

    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
    ÑĢаÑģ
    -0.16
    -за
    -0.15
    ilm
    -0.14
    zan
    -0.14
    agua
    -0.14
    udget
    -0.14
    UpInside
    -0.13
    æ´¥
    -0.13
    baum
    -0.13
    acific
    -0.13
    POSITIVE LOGITS
    ively
    0.27
    atively
    0.25
     further
    0.23
    /mit
    0.22
     from
    0.21
     us
    0.21
    ions
    0.19
    atives
    0.18
    vely
    0.17
     them
    0.17
    Activations Density 0.020%

    No Known Activations