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. 23655
    Prev
    Next
    INDEX
    Explanations

    phrases indicating potential actions or events involving community members

    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
    entes
    -0.16
    ouser
    -0.15
    ÙĨدگÛĮ
    -0.14
    «ĺ
    -0.14
    ORIES
    -0.14
     dün
    -0.13
    .getSharedPreferences
    -0.13
     voks
    -0.13
    utow
    -0.13
    _HOUR
    -0.13
    POSITIVE LOGITS
     opportunity
    1.33
     chance
    1.20
     opportunities
    1.09
     Opportunity
    1.00
    æľºä¼ļ
    0.93
     Chance
    0.90
     Opportunities
    0.89
     oportun
    0.88
     chances
    0.84
    Chance
    0.83
    Activations Density 0.047%

    No Known Activations