© 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. 22488
    Prev
    Next
    INDEX
    Explanations

    phrases related to team dynamics and communication in high-stakes situations

    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
    adol
    -0.14
     inn
    -0.14
     “
    -0.13
    atomy
    -0.13
     actual
    -0.13
     oct
    -0.13
    kok
    -0.13
     LNG
    -0.13
     epic
    -0.12
    OUCH
    -0.12
    POSITIVE LOGITS
     intel
    0.20
     briefing
    0.18
     mission
    0.17
    .assets
    0.16
    _assets
    0.16
    Intel
    0.16
     sä
    0.16
    ãģĵãģ¡ãĤī
    0.15
    à¸Ļาม
    0.15
     briefed
    0.15
    Activations Density 0.043%

    No Known Activations