© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 587
    Prev
    Next
    INDEX
    Explanations

    Online drama/harassment

    np_max-act · gemini-2.0-flash

    The main thing this neuron does is detect mentions of online harassment behavior and related fake-persona (“sock puppet”) activity.

    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
     marriages
    -0.06
    85
    -0.06
    osg
    -0.06
     notes
    -0.06
     advice
    -0.06
             
    -0.06
     scissors
    -0.06
     cards
    -0.06
    hi
    -0.06
    POSITIVE LOGITS
    (matches
    0.07
     Tro
    0.06
    .getSelectionModel
    0.06
    捕
    0.06
    ayout
    0.06
     podařilo
    0.06
     Spartan
    0.06
     περιοχή
    0.06
     chiếm
    0.06
     поврежд
    0.06
    Activations Density 0.027%

    No Known Activations