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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 54786
    Prev
    Next
    INDEX
    Explanations

    phrases indicating action or requests, particularly in formal 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
     :↵
    -0.15
    :↵
    -0.15
     aforementioned
    -0.15
     ?,
    -0.14
     freaking
    -0.14
    ’
    -0.13
    .*↵
    -0.13
    ³³
    -0.13
     ?',
    -0.13
     _______,
    -0.13
    POSITIVE LOGITS
     \↵
    0.16
    ă
    0.14
     [âĢ¦]...↵
    0.14
    \↵
    0.14
     ãħĩãħĩ
    0.14
    ÂŃtion
    0.14
     "\↵
    0.14
    ãĢģ
    0.14
    nicos
    0.14
    utterstock
    0.13
    Activations Density 10.126%

    No Known Activations