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

    Neuronpedia

    Jacobian LensNEW
    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. 38204
    Prev
    Next
    INDEX
    Explanations

    question answering

    np_max-act · gemini-2.0-flash

    The neuron detects adverbs (and related short instruction words) that specify the desired manner or style of the response (e.g. “as,” “wish,” “effectively”).

    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
     […]↵
    -0.07
    -add
    -0.05
     pozdě
    -0.05
     raining
    -0.05
    	div
    -0.05
    Anne
    -0.05
    льт
    -0.05
     bottleneck
    -0.05
    conv
    -0.05
     pop
    -0.05
    POSITIVE LOGITS
    /legal
    0.07
    .Tasks
    0.07
     COUR
    0.07
     Frequ
    0.07
     HDMI
    0.07
     JpaRepository
    0.07
     izin
    0.07
     působ
    0.07
     funkc
    0.06
    _STRUCT
    0.06
    Activations Density 0.097%

    No Known Activations