© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 54251
    Prev
    Next
    INDEX
    Explanations

    Online forum/blog text

    np_max-act · gemini-2.0-flash

    This neuron responds to structural or formatting elements—things like markup tokens, metadata markers, parentheses, edits, and other non-content “decorations” in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    direct, conversational references to participants—first- or second-person pronoun usage indicating address or personal involvement.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Variable
    -0.07
     Those
    -0.07
     предпри
    -0.07
    .with
    -0.07
    ariat
    -0.07
     Ev
    -0.06
    sam
    -0.06
    critical
    -0.06
    magnitude
    -0.06
     pitcher
    -0.06
    POSITIVE LOGITS
    ємо
    0.06
    <Key
    0.06
    ops
    0.06
    ��
    0.06
     cambi
    0.06
    .CurrentRow
    0.06
     땅
    0.06
     unpl
    0.06
    _DX
    0.06
    ]='\
    0.06
    Activations Density 0.385%

    No Known Activations