© 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. 24685
    Prev
    Next
    INDEX
    Explanations

    phrases indicating possession or existence

    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
    ानत
    -0.14
     forgot
    -0.13
    otle
    -0.13
     undergo
    -0.13
    unar
    -0.13
     полож
    -0.13
    orque
    -0.12
     threw
    -0.12
     завиÑģим
    -0.12
    ropp
    -0.12
    POSITIVE LOGITS
     been
    0.69
    been
    0.49
     Been
    0.46
     sido
    0.41
     BEEN
    0.41
    Been
    0.40
     été
    0.39
     být
    0.31
     be
    0.30
     бÑĭÑĤÑĮ
    0.30
    Activations Density 0.180%

    No Known Activations