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. 41438
    Prev
    Next
    INDEX
    Explanations

    references to examples related to societal issues, particularly discrimination and community-related policies

    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
    ivant
    -0.18
    ä»»ä½ķ
    -0.16
    enci
    -0.14
    ä¸ĢåĪĩ
    -0.14
     qualquer
    -0.14
     maduras
    -0.14
    sterol
    -0.14
    _PHP
    -0.13
    маÑĤи
    -0.13
    ainty
    -0.13
    POSITIVE LOGITS
     how
    0.66
    how
    0.49
     why
    0.47
     cómo
    0.37
    å¦Ĥä½ķ
    0.35
     HOW
    0.32
    why
    0.31
    -how
    0.30
     what
    0.29
     How
    0.28
    Activations Density 0.534%

    No Known Activations