© 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. 41326
    Prev
    Next
    INDEX
    Explanations

    references to oppression and violence in political contexts

    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
     Separ
    -0.17
    aliz
    -0.16
    lef
    -0.15
    itelist
    -0.15
    modifiable
    -0.14
    onium
    -0.14
     phóng
    -0.14
    ativity
    -0.14
     Dy
    -0.14
    izzo
    -0.14
    POSITIVE LOGITS
     repression
    0.43
     suppression
    0.41
     suppress
    0.37
     Crush
    0.37
     crush
    0.37
     crushed
    0.37
     suppressed
    0.36
     crushing
    0.35
     Crushing
    0.34
    Suppress
    0.33
    Activations Density 0.036%

    No Known Activations