© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 54168
    Prev
    Next
    INDEX
    Explanations

    prominent nouns and phrases related to significant concepts or entities

    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
    .Formatter
    -0.14
    /respond
    -0.13
    ãĤ«ãĥĨ
    -0.13
    _##
    -0.12
    _fifo
    -0.12
     inconsist
    -0.11
     Deals
    -0.11
    apesh
    -0.11
    ÑĢаÑģ
    -0.11
    noinspection
    -0.11
    POSITIVE LOGITS
     fact
    0.24
     presence
    0.21
     combination
    0.20
     mere
    0.19
     existence
    0.19
     sheer
    0.18
     absence
    0.17
     question
    0.16
    fact
    0.16
     success
    0.16
    Activations Density 0.545%

    No Known Activations