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. 23247
    Prev
    Next
    INDEX
    Explanations

    instances of the word "sign" and its variations related to endorsements and petitions

    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
    hind
    -0.15
    fork
    -0.15
    coni
    -0.14
    TED
    -0.14
    elong
    -0.14
    nels
    -0.14
    _PUSHDATA
    -0.14
    ollah
    -0.14
    vil
    -0.14
    غÙĬÙĦ
    -0.14
    POSITIVE LOGITS
     onto
    0.32
    onto
    0.28
     Ont
    0.23
    atories
    0.23
    atory
    0.21
    Ont
    0.20
     ont
    0.18
     supportive
    0.18
    ifiant
    0.18
     support
    0.17
    Activations Density 0.009%

    No Known Activations