© 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. 125155
    Prev
    Next
    INDEX
    Explanations

    apply

    np_max-act · gemini-2.0-flash

    This neuron responds to common English function words (e.g. prepositions and auxiliary verbs like “apply,” “applies,” “to,” “in,” “all”).

    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
    eração
    -0.07
    <input
    -0.06
    فق
    -0.06
     mam
    -0.06
    icit
    -0.06
    ození
    -0.06
     hacked
    -0.06
    ンダ
    -0.06
     kale
    -0.06
     món
    -0.06
    POSITIVE LOGITS
     apply
    0.10
     applies
    0.10
     indispens
    0.07
     applicable
    0.07
     exception
    0.07
     except
    0.07
     プロ
    0.07
     Apply
    0.07
     ░
    0.07
     standards
    0.07
    Activations Density 0.015%

    No Known Activations