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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 54416
    Prev
    Next
    INDEX
    Explanations

    academic and performance-related terms associated with universities and rankings

    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
    iginal
    -0.15
    frauen
    -0.15
    оÑĢи
    -0.15
    flen
    -0.15
    /*č↵
    -0.14
     зави
    -0.14
     addCriterion
    -0.14
    leck
    -0.14
    wahl
    -0.14
    jekt
    -0.14
    POSITIVE LOGITS
     ranks
    0.37
     rank
    0.36
     lead
    0.34
     leads
    0.33
     ranked
    0.31
     ranking
    0.30
     top
    0.28
     tops
    0.27
    leads
    0.27
    rank
    0.27
    Activations Density 0.139%

    No Known Activations