© 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. 54022
    Prev
    Next
    INDEX
    Explanations

    references to political affiliations and biases, particularly those aligned with conservative ideologies

    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
     progressive
    -0.29
     liberal
    -0.26
     liberalism
    -0.24
     progressives
    -0.23
     leftist
    -0.22
     Democrat
    -0.21
     Progressive
    -0.20
     socialist
    -0.20
     Democratic
    -0.19
     democr
    -0.19
    POSITIVE LOGITS
     conservative
    0.40
     conservatives
    0.39
     conserv
    0.33
     Conserv
    0.33
     right
    0.33
     Right
    0.33
     Republicans
    0.32
     conservatism
    0.32
     Conservative
    0.32
     Republican
    0.32
    Activations Density 0.079%

    No Known Activations