© 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. 25061
    Prev
    Next
    INDEX
    Explanations

    references to specific units or groups in a structured context

    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
    imuth
    -0.17
    _units
    -0.16
    hall
    -0.16
    har
    -0.16
    Units
    -0.15
    WER
    -0.15
    ลาย
    -0.15
    ëĵĿ
    -0.15
    sit
    -0.15
    sal
    -0.15
    POSITIVE LOGITS
    ary
    0.33
    errupted
    0.27
    arian
    0.25
    errupt
    0.22
    arians
    0.22
    iated
    0.21
    arity
    0.21
    asking
    0.21
    itial
    0.20
    ellig
    0.20
    Activations Density 0.007%

    No Known Activations