© 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. 26082
    Prev
    Next
    INDEX
    Explanations

    phrases indicating hypothetical or illustrative examples

    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
    omi
    -0.17
    aris
    -0.16
    éłĥ
    -0.16
    ldata
    -0.14
    PEnd
    -0.14
    agas
    -0.14
    ặn
    -0.13
    -translate
    -0.13
    éģł
    -0.13
    ترÛĮ
    -0.13
    POSITIVE LOGITS
     purposes
    0.49
     sake
    0.44
     ease
    0.35
     simplicity
    0.32
     convenience
    0.32
     purpose
    0.30
     argument
    0.27
     bre
    0.26
    pur
    0.26
    simp
    0.26
    Activations Density 0.031%

    No Known Activations