© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 63650
    Prev
    Next
    INDEX
    Explanations

    phrases related to local news and community events

    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
     Miami
    -0.16
    Miami
    -0.16
    uft
    -0.15
    avel
    -0.14
     Bilg
    -0.14
    Ñħа
    -0.14
     NYC
    -0.14
     Colbert
    -0.14
    æİĽ
    -0.13
     Brooklyn
    -0.13
    POSITIVE LOGITS
     Sacramento
    0.50
     Sac
    0.43
    Sac
    0.42
     sac
    0.34
    916
    0.29
     SAC
    0.27
     sacram
    0.27
     California
    0.24
    California
    0.23
     Sierra
    0.23
    Activations Density 0.036%

    No Known Activations