© 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. 29721
    Prev
    Next
    INDEX
    Explanations

    references to a specific individual involved in a negative 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
    uchs
    -0.16
    ISMATCH
    -0.16
    opsis
    -0.15
    lixir
    -0.14
    åĨĨ
    -0.14
     Stage
    -0.14
     menus
    -0.14
    ATCH
    -0.14
    atures
    -0.13
     meaningful
    -0.13
    POSITIVE LOGITS
    rove
    0.23
    alore
    0.20
     Madness
    0.18
    ocup
    0.17
    erial
    0.16
    als
    0.15
     Mang
    0.15
    resh
    0.15
    OLD
    0.15
    (man
    0.15
    Activations Density 0.003%

    No Known Activations