© 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. 8864
    Prev
    Next
    INDEX
    Explanations

    phrases related to coordinated public relations efforts and political engagement strategies

    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
    apel
    -0.17
    æ½
    -0.16
     Jako
    -0.16
    chos
    -0.14
     explos
    -0.13
    antu
    -0.13
    335
    -0.13
    çĪĨ
    -0.12
    404
    -0.12
    403
    -0.12
    POSITIVE LOGITS
     campaign
    0.51
     effort
    0.45
    campaign
    0.36
     campaigns
    0.35
    Campaign
    0.33
     Campaign
    0.33
     program
    0.30
     initiative
    0.30
     drive
    0.28
    ampaign
    0.28
    Activations Density 0.146%

    No Known Activations