© 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. 44862
    Prev
    Next
    INDEX
    Explanations

    references to quantities or concepts related to measurement or assessment

    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
    надлеж
    -0.16
    hiba
    -0.16
    .scalablytyped
    -0.15
    pek
    -0.14
     kür
    -0.13
    atis
    -0.13
    页éĿ¢åŃĺæ¡£å¤ĩ份
    -0.13
     bÄĻd
    -0.13
     iddi
    -0.13
    zeÅĪ
    -0.13
    POSITIVE LOGITS
    ses
    0.24
     whole
    0.17
    itis
    0.14
    adem
    0.14
     otherwise
    0.14
    osis
    0.14
     latter
    0.14
    unker
    0.13
     idea
    0.13
     numbering
    0.13
    Activations Density 0.145%

    No Known Activations