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. 23360
    Prev
    Next
    INDEX
    Explanations

    proper names, specifically those of individuals, predominantly "William" and its variations

    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
    een
    -0.17
    ORITY
    -0.17
    elles
    -0.16
    áno
    -0.15
    ôme
    -0.15
    ourg
    -0.15
    ówn
    -0.15
    ROUT
    -0.14
    ÙģÙĤ
    -0.14
    عÙĨÙĪØ§ÙĨ
    -0.14
    POSITIVE LOGITS
    boards
    0.17
    arest
    0.16
    urning
    0.16
    sons
    0.15
    ions
    0.15
    ibo
    0.15
    otte
    0.15
     Buyers
    0.15
    etus
    0.14
     Krist
    0.14
    Activations Density 0.014%

    No Known Activations