© 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 EvalsExportsinterp-engineNEWAPI 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. 24786
    Prev
    Next
    INDEX
    Explanations

    words related to emphasis and strong communication

    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
    /inet
    -0.15
    pk
    -0.15
    /from
    -0.14
    ulas
    -0.14
    .solution
    -0.14
     Antwort
    -0.14
    ahun
    -0.13
    æŃ©
    -0.13
    .rep
    -0.13
    æĬ±
    -0.13
    POSITIVE LOGITS
     importance
    0.36
     Importance
    0.30
     how
    0.25
     repeatedly
    0.21
     that
    0.20
     points
    0.19
     rằng
    0.19
     again
    0.18
     اÙĩÙħÛĮت
    0.17
    -import
    0.17
    Activations Density 0.044%

    No Known Activations