© 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. 8886
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron activates on mentions of criminal sentencing details—phrases describing being “sentenced to” prison terms, lengths of incarceration, and related judgment terminology.

    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
    _places
    -0.08
     تاریخی
    -0.07
     valuable
    -0.07
     UILabel
    -0.07
    MA
    -0.06
     altı
    -0.06
     fanatic
    -0.06
    TableView
    -0.06
    (Parse
    -0.06
     Else
    -0.06
    POSITIVE LOGITS
     lingering
    0.07
     Mans
    0.07
     threaded
    0.07
     Kirst
    0.06
    核
    0.06
    譜
    0.06
     наст
    0.06
     burgeoning
    0.06
     Yi
    0.06
    Touchable
    0.06
    Activations Density 0.019%

    No Known Activations