© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 33168
    Prev
    Next
    INDEX
    Explanations

    references to player statistics and injury updates in sports 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
    awl
    -0.16
    coles
    -0.16
    steen
    -0.15
    ÅŁa
    -0.15
    pheres
    -0.14
    ervo
    -0.14
    alfa
    -0.14
    .SC
    -0.14
    chema
    -0.14
    ÑĦекÑĤив
    -0.14
    POSITIVE LOGITS
     tight
    0.35
     wide
    0.34
     receiver
    0.33
     defensive
    0.32
     safety
    0.31
     pun
    0.30
     Tight
    0.29
     Safety
    0.29
     Wide
    0.29
     running
    0.28
    Activations Density 0.125%

    No Known Activations