© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 8656
    Prev
    Next
    INDEX
    Explanations

    parenthetical mentions and their associated details

    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
    HM
    -0.13
    .den
    -0.13
    ades
    -0.13
    cret
    -0.13
    침
    -0.13
    iž
    -0.13
     Cobb
    -0.13
    peater
    -0.12
    otel
    -0.12
    quip
    -0.12
    POSITIVE LOGITS
     pronounced
    0.42
    pron
    0.39
     pronounce
    0.38
     pron
    0.33
     Pron
    0.32
     pronunciation
    0.32
     spelled
    0.30
     spell
    0.28
    phon
    0.24
     spelling
    0.23
    Activations Density 0.054%

    No Known Activations