© 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. 44796
    Prev
    Next
    INDEX
    Explanations

    demonstrative pronouns and related phrases

    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
    ãģĹãģ¦ãģĬãĤĬ
    -0.17
    ãĤīãĤĮãģ¦ãģĦãĤĭ
    -0.17
     are
    -0.16
     spoken
    -0.16
     will
    -0.16
    will
    -0.16
     cannot
    -0.16
     is
    -0.15
     flown
    -0.15
     fallen
    -0.15
    POSITIVE LOGITS
     became
    0.43
     gave
    0.36
     began
    0.34
     took
    0.31
     grew
    0.30
     Became
    0.30
     went
    0.30
     came
    0.29
     didn
    0.27
     did
    0.26
    Activations Density 0.015%

    No Known Activations