© 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. 63312
    Prev
    Next
    INDEX
    Explanations

    sections of text that contain introductory or formatting elements, such as paragraph breaks, headers, or bullet points

    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
     PartialView
    -0.15
     but
    -0.14
    @qq
    -0.14
     cannot
    -0.13
    &id
    -0.13
     Majority
    -0.13
     only
    -0.13
     Cannot
    -0.13
     except
    -0.13
     mostly
    -0.13
    POSITIVE LOGITS
     yourselves
    0.17
     ourselves
    0.16
    atron
    0.16
    áme
    0.15
    çľī
    0.15
    ÃĹ</
    0.15
    εÏģ
    0.15
    jure
    0.15
    alama
    0.14
    -нибÑĥдÑĮ
    0.14
    Activations Density 0.017%

    No Known Activations