Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 23466
    Prev
    Next
    INDEX
    Explanations

    mentions of structural damage or destruction

    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
    relude
    -0.17
    weise
    -0.17
    pieces
    -0.16
    æł·çļĦ
    -0.16
    teen
    -0.15
    immers
    -0.15
    yal
    -0.14
    arde
    -0.14
    seed
    -0.14
    段
    -0.14
    POSITIVE LOGITS
    artment
    0.23
    ocrat
    0.21
    ÄŁi
    0.19
    beat
    0.17
    ÚĨÙĩ
    0.17
     facto
    0.17
    URIComponent
    0.15
    marshal
    0.15
    hong
    0.15
    IALOG
    0.15
    Activations Density 0.124%

    No Known Activations