Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 23805
    Prev
    Next
    INDEX
    Explanations

    phrases indicating features or characteristics of products

    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
    å®ģ
    -0.15
    tsky
    -0.14
    assen
    -0.14
    562
    -0.14
    puty
    -0.14
    aran
    -0.14
    ottage
    -0.14
    ìŀ¡ëĭ´
    -0.13
    605
    -0.13
    xae
    -0.13
    POSITIVE LOGITS
     available
    0.23
     activated
    0.21
    Available
    0.20
    available
    0.19
     present
    0.19
     triggered
    0.18
     useful
    0.18
    ienne
    0.18
     Available
    0.18
     disabled
    0.18
    Activations Density 0.065%

    No Known Activations