© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 28738
    Prev
    Next
    INDEX
    Explanations

    proper nouns associated with specific locations and landmarks

    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
     alone
    -0.18
     surrounding
    -0.17
     followed
    -0.16
     neben
    -0.16
    raft
    -0.16
    uen
    -0.16
     from
    -0.16
     besides
    -0.16
    :
    -0.16
    vis
    -0.16
    POSITIVE LOGITS
     Ùħباش
    0.27
     iParam
    0.21
    actionDate
    0.21
    Intialized
    0.18
     addCriterion
    0.18
    $MESS
    0.18
    AndHashCode
    0.18
    .scalablytyped
    0.17
    using
    0.17
    $LANG
    0.17
    Activations Density 0.045%

    No Known Activations