© 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. 25510
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a foundational or supporting principle

    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
    åĪ°
    -0.14
    straints
    -0.14
    rana
    -0.14
    à¹Ħว
    -0.14
     dab
    -0.14
    ags
    -0.13
     Dün
    -0.13
    GS
    -0.13
    ีà¹Ģà¸Ń
    -0.13
    ãĥ«ãĥĪ
    -0.13
    POSITIVE LOGITS
     on
    0.45
     upon
    0.44
    upon
    0.30
     trên
    0.30
     Upon
    0.29
     عÙĦÙī
    0.28
     auf
    0.28
     на
    0.28
     på
    0.27
     pada
    0.26
    Activations Density 0.032%

    No Known Activations