© 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 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. 42905
    Prev
    Next
    INDEX
    Explanations

    the word "as" and its variations 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
    buz
    -0.15
    avaÅŁ
    -0.15
    ignon
    -0.15
    ecd
    -0.14
    anuts
    -0.14
    iguiente
    -0.14
    ì²Ļ
    -0.14
    аÑĤков
    -0.14
    ptune
    -0.14
    asca
    -0.14
    POSITIVE LOGITS
     part
    0.37
    paragus
    0.25
    sembl
    0.25
    cert
    0.24
     an
    0.24
     follows
    0.23
    cribed
    0.23
     soon
    0.23
    sembles
    0.23
    part
    0.23
    Activations Density 0.312%

    No Known Activations