© 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. 41660
    Prev
    Next
    INDEX
    Explanations

    references to geopolitical power dynamics and superpowers

    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
    usan
    -0.16
     pů
    -0.14
     Philipp
    -0.14
    itom
    -0.14
    duk
    -0.14
    ait
    -0.14
     Tune
    -0.14
    ALES
    -0.13
    jong
    -0.13
    opo
    -0.13
    POSITIVE LOGITS
     powers
    0.61
     Powers
    0.54
    powers
    0.51
     power
    0.42
    power
    0.37
     Power
    0.35
    -power
    0.34
    Power
    0.33
    _power
    0.32
     POWER
    0.30
    Activations Density 0.077%

    No Known Activations