© 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. 24936
    Prev
    Next
    INDEX
    Explanations

    terms related to various types of work, criminal offenses, and leisure activities

    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
    _stdio
    -0.15
    iei
    -0.14
    643
    -0.13
    undle
    -0.13
     Alam
    -0.13
     approach
    -0.13
    áv
    -0.12
    ullo
    -0.12
    #!
    -0.12
     spacing
    -0.12
    POSITIVE LOGITS
     related
    0.96
    related
    0.88
    -related
    0.86
     Related
    0.81
    Related
    0.79
    _related
    0.71
     RELATED
    0.68
    .related
    0.64
    缸åħ³
    0.63
    缸éĹľ
    0.57
    Activations Density 0.083%

    No Known Activations