© 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. 29041
    Prev
    Next
    INDEX
    Explanations

    references to scientific measurements and calculations

    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
    laz
    -0.14
     cigaret
    -0.13
     Structure
    -0.13
    ofire
    -0.13
    emiz
    -0.13
    Evidence
    -0.13
    wald
    -0.13
     Wiring
    -0.13
    Ñıви
    -0.13
    iaz
    -0.13
    POSITIVE LOGITS
     values
    0.34
     tables
    0.33
    values
    0.28
     Values
    0.27
     table
    0.27
    Values
    0.26
     tab
    0.24
     VALUES
    0.23
     Tables
    0.23
    Tables
    0.23
    Activations Density 0.040%

    No Known Activations