© 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. 24856
    Prev
    Next
    INDEX
    Explanations

    negations or expressions of inability and doubt

    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
    wit
    -0.16
    ame
    -0.14
    enen
    -0.14
    vider
    -0.14
    lain
    -0.14
    .trace
    -0.14
    zap
    -0.14
    izmet
    -0.14
    ideon
    -0.14
    ospace
    -0.13
    POSITIVE LOGITS
     hardly
    0.17
    pron
    0.17
    енÑĥ
    0.16
     ever
    0.15
     ï¼İ
    0.15
     remembers
    0.15
     conform
    0.15
     minced
    0.14
     aged
    0.14
     remember
    0.14
    Activations Density 0.045%

    No Known Activations