© 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. 12485
    Prev
    Next
    INDEX
    Explanations

    instances of the article "the" or its variations to assess the frequency of its use in context

    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
     Managers
    -0.17
    buie
    -0.16
    idge
    -0.16
     Directors
    -0.16
    ÛĮÙħÛĮ
    -0.15
    idges
    -0.15
     Governors
    -0.15
    ãĥ¼ãĤ¹
    -0.15
     Presidents
    -0.15
    anship
    -0.15
    POSITIVE LOGITS
     teacher
    0.19
     hospital
    0.17
     mayor
    0.16
     president
    0.16
     professor
    0.16
     studio
    0.16
     controller
    0.16
     manufacturer
    0.16
     governor
    0.16
     coach
    0.16
    Activations Density 0.020%

    No Known Activations