Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 54600
    Prev
    Next
    INDEX
    Explanations

    players' names and their associated statistics or milestones in sports

    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
    ouro
    -0.15
    ána
    -0.14
     Dirty
    -0.13
    \Catalog
    -0.13
    fallback
    -0.13
    addock
    -0.13
    ownt
    -0.13
    иденÑĤ
    -0.13
    setIcon
    -0.12
     oven
    -0.12
    POSITIVE LOGITS
     debut
    0.53
     making
    0.53
     Making
    0.51
     make
    0.50
     makes
    0.49
    Making
    0.49
    make
    0.47
     Makes
    0.46
    making
    0.45
    makes
    0.44
    Activations Density 0.020%

    No Known Activations