© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 9167
    Prev
    Next
    INDEX
    Explanations

    topics related to votes and personal relationships

    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
    ezi
    -0.16
    ustum
    -0.16
     Got
    -0.15
    azen
    -0.15
    alah
    -0.15
    _sensitive
    -0.15
    hire
    -0.15
    tps
    -0.14
    ãĥĥãĥī
    -0.14
    ledo
    -0.14
    POSITIVE LOGITS
     added
    0.23
     removed
    0.20
     done
    0.20
     put
    0.20
     installed
    0.20
     taken
    0.19
     thrown
    0.19
     delivered
    0.18
    appen
    0.18
     handy
    0.18
    Activations Density 0.092%

    No Known Activations