© 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. 42856
    Prev
    Next
    INDEX
    Explanations

    phrases related to ease of use and convenience

    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
    ymes
    -0.15
     tast
    -0.14
    eth
    -0.14
     attent
    -0.13
    渡
    -0.13
     Mil
    -0.13
    à¹Ĥà¸Ĭ
    -0.13
     Tucker
    -0.13
     –
    -0.13
    ãģłãģª
    -0.13
    POSITIVE LOGITS
    -to
    0.31
    -use
    0.26
     use
    0.22
    Use
    0.22
     Use
    0.21
    -To
    0.20
    _use
    0.19
    .use
    0.18
    _to
    0.17
    togroup
    0.17
    Activations Density 0.048%

    No Known Activations