© 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. 43925
    Prev
    Next
    INDEX
    Explanations

    phrases related to conservation and the impact of invasive species

    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
     whe
    -0.17
    ripp
    -0.16
     Whe
    -0.16
    ãĥ¡ãĥ©
    -0.15
    ury
    -0.15
    æ´Ĺ
    -0.15
    æ¬
    -0.15
     Pis
    -0.15
     pigs
    -0.15
    kenin
    -0.14
    POSITIVE LOGITS
     traps
    0.32
     trap
    0.32
     Trap
    0.31
     trapping
    0.30
    Trap
    0.27
     steril
    0.26
     neut
    0.25
     euth
    0.23
     humane
    0.23
     trapped
    0.23
    Activations Density 0.023%

    No Known Activations