© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 15412
    Prev
    Next
    INDEX
    Explanations

    expressions of bravery and acts of heroism in military 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
    çĽĬ
    -0.15
    aan
    -0.14
    785
    -0.14
    /gtest
    -0.13
     Resistance
    -0.13
     conv
    -0.13
     vmax
    -0.13
     Revolutionary
    -0.13
    Stamped
    -0.13
    Signature
    -0.13
    POSITIVE LOGITS
     hero
    0.39
     heroes
    0.32
     heroic
    0.31
    hero
    0.30
     Hero
    0.30
    -hero
    0.29
     actions
    0.27
     saved
    0.27
     saving
    0.27
     saves
    0.27
    Activations Density 0.048%

    No Known Activations