© 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. 54132
    Prev
    Next
    INDEX
    Explanations

    references to political consulting and related activities

    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
    RuleContext
    -0.15
    xCD
    -0.14
    ivot
    -0.14
    oj
    -0.14
    FromNib
    -0.13
    ÙĬرا
    -0.13
    onec
    -0.13
    artner
    -0.13
    cope
    -0.13
    .defaults
    -0.13
    POSITIVE LOGITS
     aside
    0.82
     side
    0.81
    aside
    0.67
     Side
    0.67
    -side
    0.63
    side
    0.63
    Side
    0.60
     Aside
    0.60
     sides
    0.56
     SIDE
    0.55
    Activations Density 0.016%

    No Known Activations