© 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. 92022
    Prev
    Next
    INDEX
    Explanations

    adverbs

    np_max-act · gemini-2.0-flash

    This neuron activates on English adverbs (especially those ending in “-ly”).

    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
     one
    -0.09
     kinds
    -0.07
     One
    -0.07
     another
    -0.07
     om
    -0.07
     neighbor
    -0.07
    7
    -0.07
    &e
    -0.07
    6
    -0.07
     ups
    -0.06
    POSITIVE LOGITS
    ly
    0.12
    ally
    0.12
    aily
    0.10
    LY
    0.10
    ively
    0.09
    ALLY
    0.09
     jointly
    0.09
    Fully
    0.09
    likely
    0.09
    aly
    0.09
    Activations Density 0.558%

    No Known Activations