© 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. 44677
    Prev
    Next
    INDEX
    Explanations

    themes of destruction and division in society

    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
    eza
    -0.17
    aight
    -0.16
    eniable
    -0.15
    Å¡t
    -0.15
    elib
    -0.15
    empo
    -0.15
    ãĥ³ãĥij
    -0.14
    imer
    -0.14
    702
    -0.14
    ategorized
    -0.14
    POSITIVE LOGITS
     apart
    0.52
     Apart
    0.37
     torn
    0.33
     tear
    0.31
    Apart
    0.31
     tore
    0.30
     tearing
    0.30
     Tear
    0.28
     ripped
    0.27
    -ap
    0.25
    Activations Density 0.018%

    No Known Activations