© 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. 6762
    Prev
    Next
    INDEX
    Explanations

    Questions

    np_max-act · gemini-2.0-flash

    This neuron activates on question cues—especially on “question,” “about,” and similar tokens that mark a user asking something.

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

    whether the text is asking a question or contains question-related wording.

    oai_token-act-pair · gpt-5-miniTriggered 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
    却
    -0.07
     surprisingly
    -0.07
     Bew
    -0.07
     giản
    -0.06
    Given
    -0.06
    페이지
    -0.06
     Perspective
    -0.06
     Triangle
    -0.06
     accomplishments
    -0.06
     asserts
    -0.06
    POSITIVE LOGITS
    htub
    0.07
     پیک
    0.07
    _APB
    0.07
     Sculpt
    0.06
    _variance
    0.06
    (argc
    0.06
     ^=
    0.06
     Escorts
    0.06
     GCC
    0.06
     datasets
    0.06
    Activations Density 0.065%

    No Known Activations