© 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. 15422
    Prev
    Next
    INDEX
    Explanations

    phrases indicating peace and group cooperation

    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
     McB
    -0.17
    imli
    -0.15
    atori
    -0.14
    gewater
    -0.14
    Ðĩ
    -0.14
    ishi
    -0.14
    amac
    -0.14
    ofilm
    -0.14
    batim
    -0.13
    ilton
    -0.13
    POSITIVE LOGITS
     respect
    0.37
     Respect
    0.28
     respects
    0.27
    respect
    0.27
     respectful
    0.26
     respecting
    0.26
     respected
    0.24
    å°Ĭ
    0.22
    ä¸įä¼ļ
    0.20
     neither
    0.20
    Activations Density 0.034%

    No Known Activations