© 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. 54745
    Prev
    Next
    INDEX
    Explanations

    negations and their variations in 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
    VERRIDE
    -0.17
    æĤī
    -0.15
     поба
    -0.15
    ë²Į
    -0.15
    ÄĽtÅ¡
    -0.15
    using
    -0.14
    bons
    -0.14
    /testify
    -0.14
    åŃ£
    -0.14
    shall
    -0.14
    POSITIVE LOGITS
     alone
    0.70
    alone
    0.56
     Alone
    0.53
     solo
    0.44
    -alone
    0.40
     lone
    0.36
     Solo
    0.34
    Solo
    0.34
     solitary
    0.30
     unique
    0.29
    Activations Density 0.034%

    No Known Activations