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. 23584
    Prev
    Next
    INDEX
    Explanations

    words and phrases indicating necessity or consequences

    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
    /from
    -0.14
    ÌĢ
    -0.14
    eld
    -0.14
    ere
    -0.13
    _
    -0.13
    /etc
    -0.13
    erv
    -0.13
    eng
    -0.13
    -than
    -0.12
    ̣
    -0.12
    POSITIVE LOGITS
    .SizeType
    0.17
    /Dk
    0.16
    IIIK
    0.16
    _TAC
    0.14
    â̦↵↵↵
    0.14
    .Undef
    0.13
    .Cursors
    0.13
    â̦"↵↵
    0.13
    -LAST
    0.13
    actionDate
    0.13
    Activations Density 0.001%

    No Known Activations