© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 55820
    Prev
    Next
    INDEX
    Explanations

    analogies

    np_max-act · gemini-2.0-flash

    This neuron fires on words used in friendly, second‐person analogy explanations (e.g. “imagine,” “you,” “friends,” “have,” “want,” etc.) to signal an “imagine you…” style illustrative scenario.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    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
    Accuracy
    -0.07
    .Selection
    -0.06
     JA
    -0.06
     mantle
    -0.06
    urahan
    -0.06
    ACCOUNT
    -0.06
    .createSequentialGroup
    -0.06
    _cmds
    -0.06
    Defines
    -0.06
    usb
    -0.06
    POSITIVE LOGITS
    -grade
    0.07
     fantasies
    0.06
     grade
    0.06
    وفي
    0.06
    .slim
    0.06
    廠
    0.06
     nhiệm
    0.06
     siti
    0.06
    GRADE
    0.06
    jíž
    0.06
    Activations Density 0.032%

    No Known Activations