© 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. 54294
    Prev
    Next
    INDEX
    Explanations

    instances of the word "desaparecer" and its related forms, indicating themes of disappearance or absence

    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
    a
    -0.25
    i
    -0.22
    iag
    -0.21
    z
    -0.20
    iç
    -0.20
    y
    -0.20
    eniz
    -0.20
    e
    -0.19
    .gz
    -0.19
    aç
    -0.19
    POSITIVE LOGITS
    s
    0.42
    -vous
    0.34
    sch
    0.27
    vous
    0.27
    epam
    0.25
    gerald
    0.24
    (es
    0.22
    ska
    0.22
    ionario
    0.21
    scape
    0.21
    Activations Density 0.057%

    No Known Activations