Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 9210
    Prev
    Next
    INDEX
    Explanations

    phrases containing numbers and related quantifiers

    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
    ardon
    -0.15
    ↵↵
    -0.15
    ahun
    -0.14
    ingleton
    -0.14
    richt
    -0.14
    ÄĽk
    -0.14
    åĹ
    -0.14
    .struct
    -0.13
    uckles
    -0.13
    esco
    -0.13
    POSITIVE LOGITS
     counting
    0.46
    /or
    0.37
     Count
    0.28
    count
    0.26
    /OR
    0.26
    -count
    0.25
     still
    0.25
     count
    0.24
     Still
    0.23
     countless
    0.23
    Activations Density 0.055%

    No Known Activations