© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 43155
    Prev
    Next
    INDEX
    Explanations

    expressions of irony and emotional reflections on social and political issues

    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
    eniable
    -0.15
    icas
    -0.15
    ecess
    -0.14
    isia
    -0.14
    rowsable
    -0.13
    ='".
    -0.13
    encent
    -0.13
     Angiospermae
    -0.13
    atak
    -0.13
    ispens
    -0.13
    POSITIVE LOGITS
     iron
    0.43
     irony
    0.40
    iron
    0.39
     Iron
    0.38
    Iron
    0.37
     ironic
    0.32
    IRON
    0.25
     odd
    0.23
     interesting
    0.22
     sad
    0.20
    Activations Density 0.140%

    No Known Activations