© 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 EvalsExportsinterp-engineNEWAPI 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. 13665
    Prev
    Next
    INDEX
    Explanations

    phrases related to production and efficiency

    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
    каÑģ
    -0.16
    uru
    -0.15
    ipc
    -0.14
    าà¸Ķ
    -0.14
     Herz
    -0.14
    aku
    -0.14
    ण
    -0.13
    inue
    -0.13
    ibri
    -0.13
    ĴĮ
    -0.13
    POSITIVE LOGITS
     produce
    0.72
     producing
    0.72
     produces
    0.68
     produ
    0.62
     produced
    0.60
    produ
    0.60
     Produ
    0.58
    produce
    0.56
     Produce
    0.53
     output
    0.51
    Activations Density 0.048%

    No Known Activations