© 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. 109658
    Prev
    Next
    INDEX
    Explanations

    legal, official language

    np_max-act · gemini-2.0-flash

    This neuron activates on words in instructions describing how someone should commit illegal or immoral acts.

    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
     때
    -0.07
    umping
    -0.07
    ame
    -0.07
    AME
    -0.06
     få
    -0.06
    .pre
    -0.06
     damp
    -0.06
     ช
    -0.06
    steps
    -0.06
     aggi
    -0.06
    POSITIVE LOGITS
     refer
    0.08
    <Service
    0.08
     refers
    0.07
     Complex
    0.07
     downt
    0.07
     bra
    0.07
    통령
    0.07
    =image
    0.07
     движ
    0.07
    _chart
    0.07
    Activations Density 0.002%

    No Known Activations