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. 23706
    Prev
    Next
    INDEX
    Explanations

    references to specific characters and transformative actions in gaming and narratives

    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
    rus
    -0.18
    etary
    -0.18
    eros
    -0.17
    dle
    -0.16
    lectric
    -0.16
    zers
    -0.15
    gear
    -0.15
    hard
    -0.15
    amma
    -0.15
    lect
    -0.15
    POSITIVE LOGITS
    itect
    0.26
    edly
    0.22
    itecture
    0.21
    red
    0.19
    adia
    0.18
    pNet
    0.17
    lings
    0.16
    aic
    0.16
    'ın
    0.16
    ¶Į
    0.16
    Activations Density 0.009%

    No Known Activations