© 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. 13726
    Prev
    Next
    INDEX
    Explanations

    verbs related to actions or activities being performed

    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
    inders
    -0.16
    igers
    -0.16
    378
    -0.15
    ouch
    -0.14
    arse
    -0.14
    chn
    -0.14
    ception
    -0.14
    ิศ
    -0.14
    ohl
    -0.14
    ÄĻ
    -0.14
    POSITIVE LOGITS
     exactly
    1.33
     precisely
    1.33
     Exactly
    1.00
    Exactly
    0.90
    prec
    0.74
     exact
    0.72
     именно
    0.71
     precis
    0.71
     genau
    0.68
     Prec
    0.62
    Activations Density 0.028%

    No Known Activations