© 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. 44490
    Prev
    Next
    INDEX
    Explanations

    numerical data and references to measurements or assessments

    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
    YTE
    -0.15
    _PUSHDATA
    -0.14
    [of
    -0.13
     yOffset
    -0.13
    ãĤ¸ãĤ¢
    -0.13
    nown
    -0.13
    ãĥ¼ãĥĸ
    -0.13
    uppen
    -0.13
    SetBranch
    -0.13
    иÑĩеÑģкое
    -0.13
    POSITIVE LOGITS
     conclusion
    0.82
    Conclusion
    0.75
     Conclusion
    0.72
     summary
    0.68
     conclude
    0.67
     conclusions
    0.66
     Summary
    0.63
     concluded
    0.62
     concludes
    0.61
    summary
    0.59
    Activations Density 0.020%

    No Known Activations