© 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. 41145
    Prev
    Next
    INDEX
    Explanations

    specific character names or notable entities within 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
     with
    -0.19
    adol
    -0.16
    ToFit
    -0.15
    with
    -0.15
     vỼi
    -0.15
    swith
    -0.15
     dengan
    -0.14
     withString
    -0.14
     withObject
    -0.14
    ">//
    -0.13
    POSITIVE LOGITS
     as
    0.28
     involved
    0.25
     sebagai
    0.24
    ä½ľä¸º
    0.23
     present
    0.23
     included
    0.20
     Included
    0.20
     als
    0.19
     каÑĩеÑģÑĤве
    0.19
     Present
    0.19
    Activations Density 0.057%

    No Known Activations