© 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. 22310
    Prev
    Next
    INDEX
    Explanations

    phrases related to external demands and disruptions in established 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
    antz
    -0.16
    dlg
    -0.15
     trapping
    -0.15
     humiliating
    -0.15
     trap
    -0.15
    кÑĤи
    -0.15
    eya
    -0.15
    éĥİ
    -0.15
    datagrid
    -0.15
     Trap
    -0.14
    POSITIVE LOGITS
     pert
    0.40
     disrupt
    0.40
     disturbance
    0.38
     disturb
    0.38
    pert
    0.36
     disruption
    0.35
     Pert
    0.35
     distur
    0.32
     upset
    0.32
     disrupting
    0.32
    Activations Density 0.054%

    No Known Activations