© 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
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 30536
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal issues and charges

    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
    лÑıÑħ
    -0.15
     nackte
    -0.14
    δÎŃ
    -0.14
    ãĤĩ
    -0.14
    unik
    -0.13
    ptrdiff
    -0.13
    oust
    -0.12
    dent
    -0.12
    locale
    -0.12
    rene
    -0.12
    POSITIVE LOGITS
     multiple
    1.14
     Multiple
    0.94
    multiple
    0.92
    Multiple
    0.88
    _multiple
    0.75
    ultiple
    0.72
    å¤ļ
    0.71
     several
    0.65
     ìŬ룬
    0.55
     å¤ļ
    0.54
    Activations Density 0.114%

    No Known Activations