© 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 EvalsExportsinterp-engineNEWAPI 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. 43836
    Prev
    Next
    INDEX
    Explanations

    references to deception or imitation

    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
    iyon
    -0.15
    isses
    -0.14
    lassen
    -0.14
    andr
    -0.14
    éĶģ
    -0.14
     fkk
    -0.13
    éİ
    -0.13
    ÑĢовиÑĩ
    -0.13
    illance
    -0.13
    tuk
    -0.13
    POSITIVE LOGITS
     holding
    0.52
     hold
    0.48
     Holding
    0.43
     HOLD
    0.42
     Hold
    0.41
     holds
    0.40
    holding
    0.40
    hold
    0.39
    Hold
    0.39
     held
    0.37
    Activations Density 0.097%

    No Known Activations