© 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. 33413
    Prev
    Next
    INDEX
    Explanations

    phrases indicating potential or opportunity

    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
     Bucc
    -0.16
    MB
    -0.15
     somehow
    -0.15
    375
    -0.15
    hey
    -0.15
    QC
    -0.14
    oram
    -0.14
    à¸Ľà¸£à¸°à¸§
    -0.14
    ersive
    -0.14
    piece
    -0.13
    POSITIVE LOGITS
     else
    0.22
    /stdc
    0.18
     ELSE
    0.18
    rido
    0.18
    ByExample
    0.17
    Else
    0.17
    	else
    0.17
     wrong
    0.16
    eno
    0.16
     besides
    0.15
    Activations Density 0.044%

    No Known Activations