© 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. 41040
    Prev
    Next
    INDEX
    Explanations

    mentions of sports players and teams

    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
    iw
    -0.15
    ially
    -0.15
     Yao
    -0.15
     tung
    -0.14
    Borders
    -0.14
     Summers
    -0.14
    ignum
    -0.14
    iken
    -0.14
    ancock
    -0.13
     Eag
    -0.13
    POSITIVE LOGITS
     Atlanta
    0.18
     Crew
    0.17
     MLS
    0.16
    loi
    0.16
    FC
    0.16
    isine
    0.16
     Sporting
    0.16
    atürk
    0.15
    Atlanta
    0.15
     Montreal
    0.15
    Activations Density 0.032%

    No Known Activations