© 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. 30091
    Prev
    Next
    INDEX
    Explanations

    expressions relating to lifestyle and modes of living

    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
    ãĥ³ãĥij
    -0.18
    culate
    -0.15
    uos
    -0.15
    ãģ£ãģį
    -0.15
    rou
    -0.14
    ulp
    -0.14
    Disposition
    -0.14
    irt
    -0.14
    oup
    -0.14
     sources
    -0.13
    POSITIVE LOGITS
     life
    0.25
     sorts
    0.20
     vida
    0.19
    life
    0.19
     thought
    0.17
    Life
    0.17
     attack
    0.17
    operation
    0.17
    icers
    0.16
     ↵↵
    0.16
    Activations Density 0.127%

    No Known Activations