© 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. 25139
    Prev
    Next
    INDEX
    Explanations

    punctuation and conjunctions indicating relationships between ideas

    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
    olk
    -0.15
    олÑĮно
    -0.13
     attempt
    -0.13
     forIndexPath
    -0.13
    erras
    -0.13
    ibur
    -0.12
     attempted
    -0.12
    older
    -0.12
    adero
    -0.12
    ssc
    -0.12
    POSITIVE LOGITS
     having
    0.83
    having
    0.69
     Having
    0.67
    Having
    0.63
     ayant
    0.40
     knowing
    0.38
     memiliki
    0.34
     avoir
    0.34
     presence
    0.32
     possessing
    0.29
    Activations Density 0.110%

    No Known Activations