© 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. 42748
    Prev
    Next
    INDEX
    Explanations

    phrases that express approximation or nearness in comparison

    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
     exactly
    -0.22
     Exactly
    -0.18
    Exactly
    -0.15
    ients
    -0.15
     very
    -0.15
     much
    -0.15
     Much
    -0.14
    iÄįe
    -0.14
    ãĤĪãĤĬ
    -0.14
     precisely
    -0.14
    POSITIVE LOGITS
     as
    0.23
     à¤ĩतन
    0.20
     nÃło
    0.19
    degree
    0.18
     enough
    0.18
     anywhere
    0.16
    åĨĬ
    0.16
    Degree
    0.15
     nor
    0.15
    à¹Ħหà¸Ļ
    0.15
    Activations Density 0.017%

    No Known Activations