© 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. 15656
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate a main point or argument being presented

    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
    anela
    -0.17
    imits
    -0.16
    iders
    -0.16
    ستÙĩ
    -0.15
    pe
    -0.15
    ael
    -0.14
     points
    -0.14
    atten
    -0.14
    ippet
    -0.14
    ack
    -0.14
    POSITIVE LOGITS
     being
    0.19
    being
    0.17
     here
    0.17
    à¸Ĥà¸Ńà¸ĩà¸ľ
    0.17
    çĦ¶
    0.15
     Being
    0.15
    .ov
    0.15
    ůst
    0.14
    ghest
    0.14
    Being
    0.14
    Activations Density 0.006%

    No Known Activations