© 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. 40836
    Prev
    Next
    INDEX
    Explanations

    statistical data regarding voting patterns and margins

    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
    ÏĥÏĦά
    -0.16
    evin
    -0.15
    olib
    -0.14
    oman
    -0.14
    Äı
    -0.13
    à¥įपन
    -0.13
    Leod
    -0.13
     Serializable
    -0.13
     singled
    -0.13
    ваÑĢ
    -0.13
    POSITIVE LOGITS
     margin
    0.42
     margins
    0.39
    margin
    0.34
     Margin
    0.34
    Margin
    0.33
    -margin
    0.29
     marg
    0.28
     Marg
    0.26
     overwhelming
    0.25
    	margin
    0.23
    Activations Density 0.029%

    No Known Activations