© 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. 24918
    Prev
    Next
    INDEX
    Explanations

    actions and movements related to going and falling

    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
    icana
    -0.16
    ufen
    -0.16
     aby
    -0.15
    дж
    -0.14
    ilen
    -0.14
     einz
    -0.14
    iÄį
    -0.14
    upo
    -0.14
    orex
    -0.13
    ë©
    -0.13
    POSITIVE LOGITS
    -to
    0.39
    åĪ°
    0.35
     unto
    0.32
     to
    0.31
     Ø¥ÙĦÙī
    0.31
     åĪ°
    0.29
     zur
    0.28
     اÙĦÙī
    0.26
    åĪ°äºĨ
    0.26
    åĪ°çļĦ
    0.25
    Activations Density 0.031%

    No Known Activations