© 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
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 24299
    Prev
    Next
    INDEX
    Explanations

    phrases indicating numerical comparisons or quantifications

    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
    oug
    -0.15
     Nearly
    -0.14
     Nin
    -0.14
     nineteenth
    -0.14
     Almost
    -0.14
    ocoa
    -0.14
     Eleven
    -0.14
    ãĤº
    -0.13
    infra
    -0.13
    etsk
    -0.13
    POSITIVE LOGITS
    100
    0.50
    500
    0.45
    400
    0.42
    300
    0.41
    250
    0.37
    50
    0.37
    150
    0.36
    600
    0.36
    700
    0.34
    800
    0.34
    Activations Density 0.127%

    No Known Activations