© 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. 109768
    Prev
    Next
    INDEX
    Explanations

    sports and performance

    np_max-act · gemini-2.0-flash

    The neuron fires on verbs naming a person’s core professional activity (e.g. “wrestled,” “sang,” “racing,” “sailing”).

    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
     alliance
    -0.07
     adding
    -0.07
    builders
    -0.07
     organiz
    -0.07
     builder
    -0.07
     Solve
    -0.06
     Address
    -0.06
    .Results
    -0.06
     organis
    -0.06
     models
    -0.06
    POSITIVE LOGITS
    (gulp
    0.06
     irrig
    0.06
     Brenda
    0.06
    NM
    0.06
    poi
    0.06
    ùa
    0.06
     행복
    0.06
    leave
    0.06
    Vectors
    0.06
    させ
    0.05
    Activations Density 0.086%

    No Known Activations