© 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. 22249
    Prev
    Next
    INDEX
    Explanations

    specific measurements and details related to physical objects or equipment, especially in instructional contexts

    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
    meter
    -0.16
    oku
    -0.16
    uru
    -0.15
    htub
    -0.15
    ater
    -0.15
    uir
    -0.15
    ITU
    -0.15
    ional
    -0.15
    109
    -0.14
    -meter
    -0.14
    POSITIVE LOGITS
     inches
    0.43
     inch
    0.37
     Inches
    0.33
    inch
    0.30
     Inch
    0.29
    -inch
    0.29
     foot
    0.21
     ounces
    0.21
    _inches
    0.20
     اÛĮÙĨÚĨ
    0.18
    Activations Density 0.032%

    No Known Activations