© 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. 41631
    Prev
    Next
    INDEX
    Explanations

    connections related to alliances and betrayals in a narrative context

    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
    olon
    -0.16
    roje
    -0.15
    erca
    -0.14
    bourg
    -0.14
    duc
    -0.14
    رÙĤ
    -0.14
     sợ
    -0.13
    ç¿
    -0.13
    asure
    -0.13
    .jav
    -0.13
    POSITIVE LOGITS
     align
    0.31
     siding
    0.30
     support
    0.30
     allied
    0.30
     aligned
    0.29
     alliance
    0.28
     ally
    0.27
     sided
    0.26
     neutral
    0.26
     alignment
    0.25
    Activations Density 0.093%

    No Known Activations