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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 40739
    Prev
    Next
    INDEX
    Explanations

    technical terms related to image processing and data structures

    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
     restaurant
    -0.20
     rivers
    -0.20
     restaurants
    -0.19
     rifles
    -0.19
    _radius
    -0.19
     radius
    -0.18
     रह
    -0.18
     ÑĢайонÑĥ
    -0.17
     routers
    -0.17
    .runtime
    -0.17
    POSITIVE LOGITS
     unregister
    0.22
    ru
    0.20
    RR
    0.18
    _unregister
    0.18
    .unregister
    0.17
    ueva
    0.16
     Lilly
    0.16
     unre
    0.16
     Ru
    0.15
    ÑĢÑĥ
    0.15
    Activations Density 0.031%

    No Known Activations