© 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. 32978
    Prev
    Next
    INDEX
    Explanations

    phrases related to political campaigning and advertisements

    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
     YÃĸ
    -0.15
    bbe
    -0.15
    prene
    -0.14
    phet
    -0.13
    pageTitle
    -0.13
    ../../../../
    -0.13
    steder
    -0.13
    ddit
    -0.13
    athers
    -0.13
    ibly
    -0.13
    POSITIVE LOGITS
     ads
    0.32
     ad
    0.28
     attack
    0.27
     commercials
    0.25
     Ads
    0.24
     attacking
    0.23
    attack
    0.22
    Ads
    0.22
     rob
    0.22
     negative
    0.21
    Activations Density 0.074%

    No Known Activations