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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 44251
    Prev
    Next
    INDEX
    Explanations

    references to structural damage or collapse of buildings

    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
    ิà¸ķย
    -0.15
     Cemetery
    -0.15
    //{{
    -0.14
    442
    -0.14
    æijĨ
    -0.14
    ÃŃt
    -0.14
    era
    -0.14
    otten
    -0.13
    \Active
    -0.13
     пÑĢож
    -0.13
    POSITIVE LOGITS
     building
    0.84
     buildings
    0.78
     Building
    0.65
    building
    0.65
     Buildings
    0.62
    Building
    0.60
    -building
    0.57
     BUILD
    0.54
    _building
    0.50
    build
    0.49
    Activations Density 0.098%

    No Known Activations