© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 25460
    Prev
    Next
    INDEX
    Explanations

    references to social issues and crises, particularly related to food and health

    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
    olit
    -0.17
    erez
    -0.16
    enis
    -0.15
    ادÙħ
    -0.14
    unch
    -0.14
    [float
    -0.14
     alone
    -0.13
    -haspopup
    -0.13
    endid
    -0.13
     Alone
    -0.13
    POSITIVE LOGITS
     actually
    1.09
    actually
    0.91
     Actually
    0.88
    Actually
    0.79
     actual
    0.70
    åħ¶å®ŀ
    0.63
     aslında
    0.62
    å®ŀéĻħ
    0.61
     really
    0.54
    actual
    0.53
    Activations Density 0.082%

    No Known Activations