© 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. 24298
    Prev
    Next
    INDEX
    Explanations

    phrases related to the denial or rejection of external influences or interventions

    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
    ÑĢеменно
    -0.17
    jen
    -0.15
     kabil
    -0.14
    æk
    -0.14
    ético
    -0.14
    geç
    -0.14
     namoro
    -0.14
    олож
    -0.13
    à¸ļาà¸ĩ
    -0.13
     Incontri
    -0.13
    POSITIVE LOGITS
     any
    0.82
    any
    0.55
    ä»»ä½ķ
    0.52
     Any
    0.51
     ANY
    0.51
    Any
    0.49
    .any
    0.44
    -any
    0.43
     anything
    0.43
     qualquer
    0.42
    Activations Density 0.085%

    No Known Activations