© 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. 92553
    Prev
    Next
    INDEX
    Explanations

    devotion

    np_max-act · gemini-2.0-flash

    The neuron fires on words expressing strong personal commitment or devotion (e.g. “devoted”).

    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
     ladder
    -0.08
     gamble
    -0.07
     risky
    -0.07
     lump
    -0.07
     rủi
    -0.07
     challenges
    -0.07
    >↵↵↵↵↵
    -0.07
    ULD
    -0.06
     правила
    -0.06
     Zhang
    -0.06
    POSITIVE LOGITS
     devoted
    0.10
     devotion
    0.09
    ov
    0.07
     dominating
    0.07
     WW
    0.07
     devour
    0.07
    .moveToFirst
    0.07
    او
    0.07
    OV
    0.07
     Vol
    0.07
    Activations Density 0.003%

    No Known Activations