© 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. 28114
    Prev
    Next
    INDEX
    Explanations

    instances of perceived deception or unfair treatment

    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
     ctor
    -0.15
    alim
    -0.15
    รà¸ĵ
    -0.15
     Sink
    -0.15
    457
    -0.14
    ellig
    -0.14
    cek
    -0.14
    otel
    -0.14
    ound
    -0.14
    ocator
    -0.14
    POSITIVE LOGITS
     shaft
    0.26
     hood
    0.26
     fle
    0.24
     betrayed
    0.24
     exploited
    0.23
    stab
    0.22
     cheated
    0.22
     Hood
    0.21
     betrayal
    0.20
     dup
    0.20
    Activations Density 0.032%

    No Known Activations