© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 28443
    Prev
    Next
    INDEX
    Explanations

    phrases related to significant transitions or changes

    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
    inski
    -0.16
    Æ¡i
    -0.15
    antz
    -0.15
     timing
    -0.15
     timely
    -0.14
    ÑĥÑĪки
    -0.14
    timing
    -0.14
    idlo
    -0.14
     Timing
    -0.14
    etri
    -0.14
    POSITIVE LOGITS
     era
    0.80
     age
    0.72
     Era
    0.71
     Age
    0.68
    Age
    0.55
    era
    0.53
    -era
    0.47
     AGE
    0.47
    æĹ¶ä»£
    0.46
     ERA
    0.46
    Activations Density 0.049%

    No Known Activations