© 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. 30540
    Prev
    Next
    INDEX
    Explanations

    expressing uncertainty

    np_max-act · gemini-2.0-flash

    The neuron primarily activates on special control/formatting tokens (e.g. header delimiters and end-of-text markers) rather than ordinary words.

    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
    _mgmt
    -0.08
     конце
    -0.07
     ケ
    -0.07
     aroma
    -0.06
     полі
    -0.06
     csak
    -0.06
    、マ
    -0.06
     学
    -0.06
     лют
    -0.06
    Joe
    -0.06
    POSITIVE LOGITS
     oral
    0.07
     permanent
    0.06
     decis
    0.06
    Administration
    0.06
     imageSize
    0.06
    roman
    0.06
     financed
    0.06
     LNG
    0.06
     اشاره
    0.06
     мет
    0.06
    Activations Density 0.099%

    No Known Activations