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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 23242
    Prev
    Next
    INDEX
    Explanations

    the preposition "at"

    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
    AsStream
    -0.14
     setSupportActionBar
    -0.14
    erras
    -0.14
    ÎŃÏģγ
    -0.13
    levant
    -0.13
    ики
    -0.13
    à¹Įà¸Ķ
    -0.13
    utzer
    -0.13
    udev
    -0.13
    abras
    -0.13
    POSITIVE LOGITS
     any
    0.24
     anytime
    0.24
    Any
    0.17
     Any
    0.16
    any
    0.16
     bất
    0.16
    oyer
    0.16
    ä»»ä½ķ
    0.15
     ANY
    0.14
    iyah
    0.14
    Activations Density 0.003%

    No Known Activations