© 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. 42220
    Prev
    Next
    INDEX
    Explanations

    phrases related to political conspiracy and government involvement in 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
     è³
    -0.16
    ihan
    -0.15
    ida
    -0.14
     Destructor
    -0.14
     menacing
    -0.13
     Picasso
    -0.13
    áŁ
    -0.13
    azard
    -0.13
     chilling
    -0.13
    bia
    -0.13
    POSITIVE LOGITS
     staged
    0.24
     Moss
    0.22
     staging
    0.19
     moss
    0.19
    stage
    0.19
    åIJ¹
    0.18
    llum
    0.17
     False
    0.17
     Agenda
    0.17
    -stage
    0.16
    Activations Density 0.042%

    No Known Activations