© 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. 12235
    Prev
    Next
    INDEX
    Explanations

    phrases indicating annual events or activities related to celebrations or significant occurrences

    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
     Evening
    -0.16
     daytime
    -0.15
    utto
    -0.15
     afternoon
    -0.14
     nighttime
    -0.14
     evening
    -0.14
    opr
    -0.14
     Days
    -0.14
    iko
    -0.14
    alin
    -0.13
    POSITIVE LOGITS
     every
    0.90
     each
    0.77
    æ¯ı
    0.75
    every
    0.71
     Every
    0.70
    each
    0.66
    Every
    0.65
     chaque
    0.65
     æ¯ı
    0.62
     má»Ĺi
    0.61
    Activations Density 0.050%

    No Known Activations