© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 30414
    Prev
    Next
    INDEX
    Explanations

    references to legal cases and government actions related to abortion rights and healthcare policies

    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
    _VC
    -0.16
    -basket
    -0.15
    sten
    -0.15
    annel
    -0.14
    hear
    -0.14
    zano
    -0.14
     adjud
    -0.13
    engeance
    -0.13
     Plantae
    -0.13
    usz
    -0.13
    POSITIVE LOGITS
     argument
    0.35
     arguments
    0.34
     argue
    0.32
     Argument
    0.32
     argued
    0.31
     arguing
    0.30
    argument
    0.29
     args
    0.29
    Argument
    0.28
     argues
    0.28
    Activations Density 0.041%

    No Known Activations