© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 13118
    Prev
    Next
    INDEX
    Explanations

    references to data sampling and analytics issues in the context of web traffic

    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
    TestCategory
    -0.13
    HashTable
    -0.13
    DBObject
    -0.12
    ุà¸Ļ
    -0.12
    dyn
    -0.11
    NodeType
    -0.11
    itty
    -0.11
    dent
    -0.11
    opsis
    -0.11
    iko
    -0.11
    POSITIVE LOGITS
     data
    1.19
     Data
    0.98
    data
    0.94
    -data
    0.89
    _data
    0.88
    Data
    0.88
    æķ°æį®
    0.85
    .data
    0.84
     DATA
    0.83
    (data
    0.77
    Activations Density 0.018%

    No Known Activations