© 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. 55092
    Prev
    Next
    INDEX
    Explanations

    the word "of" and its variations appearing in different contexts

    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
    .scalablytyped
    -0.24
    ä¸ĬãģĮ
    -0.14
    以为
    -0.13
    Ïĩεία
    -0.13
    scribe
    -0.13
    .ONE
    -0.13
    foon
    -0.13
    /tcp
    -0.13
    ãĤ¤ãĥ¤
    -0.13
    podob
    -0.13
    POSITIVE LOGITS
     few
    0.40
     those
    0.35
     several
    0.34
     only
    0.29
     handful
    0.28
     many
    0.27
     Those
    0.26
    few
    0.24
     Several
    0.24
    Few
    0.24
    Activations Density 0.054%

    No Known Activations