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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 41625
    Prev
    Next
    INDEX
    Explanations

    topics related to government censorship and right-wing extremism

    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
    εÏī
    -0.17
    ohana
    -0.17
    ApplicationBuilder
    -0.16
    обов
    -0.16
     âĪĢ
    -0.16
    esome
    -0.16
    mé
    -0.14
    .Generated
    -0.14
    abbo
    -0.14
     αÏĥÏĦ
    -0.14
    POSITIVE LOGITS
     Greek
    0.75
     Greece
    0.69
    Greek
    0.62
     Athens
    0.62
     Greeks
    0.61
     Athen
    0.46
    GREE
    0.38
     гÑĢ
    0.38
     Gree
    0.36
     гÑĢе
    0.36
    Activations Density 0.023%

    No Known Activations