© 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. 13544
    Prev
    Next
    INDEX
    Explanations

    phrases related to durability and reliability

    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
    ยà¸Ńà¸Ķ
    -0.15
    uset
    -0.15
    outu
    -0.14
    547
    -0.14
     tamam
    -0.14
    -REAL
    -0.13
    oste
    -0.13
    :selected
    -0.13
    기ìĹIJ
    -0.13
    ulle
    -0.13
    POSITIVE LOGITS
     hold
    0.78
     holds
    0.74
     holding
    0.67
     Hold
    0.65
     Holds
    0.63
    Hold
    0.61
    hold
    0.60
     HOLD
    0.58
     held
    0.57
    holds
    0.57
    Activations Density 0.044%

    No Known Activations