© 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. 30070
    Prev
    Next
    INDEX
    Explanations

    references to the influences and consequences of media fragmentation and biases

    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
    .googlecode
    -0.16
    ardown
    -0.16
     phúc
    -0.14
     impro
    -0.14
    ÄIJT
    -0.14
    ulan
    -0.13
    Insensitive
    -0.13
     bureaucr
    -0.13
     setBackgroundImage
    -0.13
    Pdf
    -0.13
    POSITIVE LOGITS
     media
    0.40
     news
    0.39
     Media
    0.35
    media
    0.32
    -media
    0.31
    Media
    0.31
     News
    0.30
    /news
    0.29
    news
    0.28
     MEDIA
    0.27
    Activations Density 0.071%

    No Known Activations