© 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. 41198
    Prev
    Next
    INDEX
    Explanations

    expressions related to emotional depth and personal qualities

    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
    .scalablytyped
    -0.15
    iasi
    -0.15
    :CGRect
    -0.15
    odef
    -0.14
    Ø°
    -0.14
    -Saharan
    -0.14
    CommandLine
    -0.14
    ůl
    -0.14
    ushing
    -0.14
    ÌĨ
    -0.13
    POSITIVE LOGITS
     needs
    0.78
     must
    0.78
     need
    0.75
     Must
    0.67
    needs
    0.66
    must
    0.66
    need
    0.65
     Needs
    0.63
     MUST
    0.62
    Must
    0.61
    Activations Density 0.064%

    No Known Activations