© 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. 22862
    Prev
    Next
    INDEX
    Explanations

    phrases related to economic indicators and consumer spending

    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
    orra
    -0.17
    -event
    -0.14
    geme
    -0.14
    èo
    -0.14
     event
    -0.14
     zab
    -0.14
    ITICAL
    -0.14
    оÑĢод
    -0.14
    aku
    -0.14
    à¸Ļว
    -0.14
    POSITIVE LOGITS
     purchasing
    0.47
     real
    0.36
     Purch
    0.34
     buying
    0.33
     Real
    0.31
     REAL
    0.31
     Buying
    0.29
    real
    0.29
     disposable
    0.28
    (real
    0.27
    Activations Density 0.057%

    No Known Activations