© 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. 15611
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the presence of evidence or research findings

    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
    ë²Ī
    -0.15
     внеÑģ
    -0.14
     ÑĢаÑħ
    -0.13
    isse
    -0.13
    andro
    -0.13
    obuf
    -0.13
    defs
    -0.13
    odom
    -0.13
     guaranteed
    -0.13
    adol
    -0.12
    POSITIVE LOGITS
     strong
    0.45
     evidence
    0.39
    strong
    0.35
     Strong
    0.34
    Strong
    0.33
     compelling
    0.33
     strongest
    0.31
     Evidence
    0.31
     solid
    0.30
     circum
    0.29
    Activations Density 0.134%

    No Known Activations