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

    Neuronpedia

    Jacobian LensNEW
    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. 42849
    Prev
    Next
    INDEX
    Explanations

    references to specific groups or entities in discussions

    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
     was
    -0.27
     will
    -0.27
    Was
    -0.27
     wasn
    -0.26
    æľĥ
    -0.25
    will
    -0.25
    was
    -0.25
     Was
    -0.24
     sẽ
    -0.24
     were
    -0.24
    POSITIVE LOGITS
     continue
    0.25
     continues
    0.24
    contin
    0.19
    continue
    0.19
     has
    0.17
     continu
    0.17
     tend
    0.17
     telah
    0.17
    _continue
    0.16
     tends
    0.16
    Activations Density 0.010%

    No Known Activations