© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 33032
    Prev
    Next
    INDEX
    Explanations

    instances of explanation or description of methods and their impacts

    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
     anymore
    -0.15
    oine
    -0.14
    è¡¡
    -0.14
    psz
    -0.14
     вдÑĢÑĥг
    -0.14
    ieri
    -0.14
    idden
    -0.14
    ometr
    -0.14
    orie
    -0.13
    ãģĿ
    -0.13
    POSITIVE LOGITS
     thus
    0.89
    thus
    0.78
     Thus
    0.67
    Thus
    0.67
     asÃŃ
    0.50
     ainsi
    0.50
     così
    0.48
    è¿Ļæł·
    0.38
     thereby
    0.34
     ìĿ´ëłĩê²Į
    0.33
    Activations Density 0.037%

    No Known Activations