© 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. 63253
    Prev
    Next
    INDEX
    Explanations

    conditional phrases or arguments in discussions about decision-making

    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
     still
    -0.21
    still
    -0.18
     yes
    -0.18
     indeed
    -0.18
    Still
    -0.17
     we
    -0.17
     Still
    -0.17
    Yes
    -0.16
     just
    -0.16
     STILL
    -0.16
    POSITIVE LOGITS
     presently
    0.20
     unless
    0.19
     currently
    0.19
    unless
    0.17
     until
    0.17
    Currently
    0.17
     Until
    0.17
     Currently
    0.17
     increasingly
    0.17
    currently
    0.16
    Activations Density 0.002%

    No Known Activations