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. 23380
    Prev
    Next
    INDEX
    Explanations

    references to rinsing and related processes

    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
    es
    -0.28
    al
    -0.24
    ed
    -0.23
    ar
    -0.20
    i
    -0.20
    esz
    -0.20
    esco
    -0.20
    Ñĥ
    -0.20
    a
    -0.18
    iya
    -0.18
    POSITIVE LOGITS
    ny
    0.22
    shaw
    0.20
    ucle
    0.20
    omial
    0.19
    cola
    0.19
    ners
    0.19
    dependent
    0.18
    ns
    0.18
    sha
    0.18
    ews
    0.18
    Activations Density 0.057%

    No Known Activations