© 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. 42628
    Prev
    Next
    INDEX
    Explanations

    phrases indicating escalation or a significant increase in intensity or level

    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
    ece
    -0.15
    lepÅ¡ÃŃ
    -0.14
    overy
    -0.14
    ëŀij
    -0.14
    @testable
    -0.13
    processable
    -0.13
    à¹ģà¸ļ
    -0.13
    ERIC
    -0.13
    xbe
    -0.13
    à¸Ľà¸£à¸°à¸Īำ
    -0.13
    POSITIVE LOGITS
     new
    0.53
     whole
    0.42
    new
    0.40
    (new
    0.36
    whole
    0.36
     Whole
    0.34
    æĸ°çļĦ
    0.34
    -new
    0.34
    =new
    0.34
    à¹ĥหม
    0.33
    Activations Density 0.022%

    No Known Activations