© 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. 13759
    Prev
    Next
    INDEX
    Explanations

    specific nouns and terms related to individuals, places, and various contexts, especially in news or historical narratives

    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
    $MESS
    -0.19
    ¶ģ
    -0.17
    ¯¼
    -0.16
    «ĺ
    -0.16
    NewProp
    -0.16
     withObject
    -0.15
    MemoryWarning
    -0.15
    GuidId
    -0.15
    $fdata
    -0.15
    //{{
    -0.14
    POSITIVE LOGITS
    -turned
    0.19
     (
    0.14
    ãĢģ
    0.14
    ķ
    0.13
     ceremon
    0.13
     blah
    0.13
     romant
    0.13
    -era
    0.12
    Âł
    0.12
     trag
    0.12
    Activations Density 0.582%

    No Known Activations