© 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. 13353
    Prev
    Next
    INDEX
    Explanations

    the word "became" as a key indicator of change or transformation

    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
    âĢ¦↵
    -0.16
    _deinit
    -0.14
     âĢ¦↵
    -0.14
    SSERT
    -0.14
    okino
    -0.14
    atos
    -0.14
    dealloc
    -0.13
    ubat
    -0.13
    .LayoutStyle
    -0.12
    achinery
    -0.12
    POSITIVE LOGITS
    ometr
    0.17
    uard
    0.14
    exampleInputEmail
    0.14
    reau
    0.14
    xt
    0.13
    loor
    0.12
    ÅĤa
    0.12
    ooke
    0.12
    ÐŁÐļ
    0.12
    rá
    0.12
    Activations Density 0.000%

    No Known Activations