© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 45162
    Prev
    Next
    INDEX
    Explanations

    references to life, death, and occurrences surrounding mortality

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
     Davidson
    -0.19
     Dagger
    -0.17
    inium
    -0.17
     Dunn
    -0.16
     deltas
    -0.15
    adu
    -0.15
    etten
    -0.15
     Dos
    -0.15
    Dos
    -0.14
    VOKE
    -0.14
    POSITIVE LOGITS
     death
    0.71
    death
    0.57
     Death
    0.56
    -death
    0.54
    Death
    0.52
     DE
    0.48
     deaths
    0.46
    _death
    0.43
     æŃ
    0.38
     ÑģмеÑĢÑĤи
    0.38
    Activations Density 0.064%

    No Known Activations