© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 45269
    Prev
    Next
    INDEX
    Explanations

    references to specific brands or companies involved in customer service situations

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    ÙĨÚ¯ÛĮ
    -0.16
    utsche
    -0.15
     Dana
    -0.15
    anium
    -0.14
     storefront
    -0.14
    اÙĬت
    -0.14
    ALT
    -0.14
    dana
    -0.13
    ноÑģÑĤ
    -0.13
     Stout
    -0.13
    POSITIVE LOGITS
     UK
    0.19
     £
    0.19
     Welsh
    0.18
     specification
    0.17
    ä»
    0.16
     Specification
    0.16
     Britain
    0.16
     WL
    0.16
    eni
    0.15
    Specification
    0.15
    Activations Density 0.009%

    No Known Activations