© 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. 41120
    Prev
    Next
    INDEX
    Explanations

    references to "par" related terms that imply categories or classifications

    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
    emas
    -0.19
    ril
    -0.16
    -0.16
    ene
    -0.16
    egal
    -0.16
    braco
    -0.16
    estate
    -0.16
    elier
    -0.15
    lyph
    -0.15
    дÑĸл
    -0.15
    POSITIVE LOGITS
     excellence
    0.22
    adox
    0.22
    abolic
    0.21
    sons
    0.21
    odies
    0.20
    liament
    0.19
    lor
    0.19
    agon
    0.18
    ox
    0.18
    son
    0.18
    Activations Density 0.016%

    No Known Activations