© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 54528
    Prev
    Next
    INDEX
    Explanations

    Negative emotions

    np_max-act · gemini-2.0-flash

    The neuron is detecting words and phrases related to self-harm, suicidal thoughts, and emotional distress.

    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
    ушка
    -0.07
    uc
    -0.06
     /\
    -0.06
    '><
    -0.06
     визначення
    -0.06
    戒
    -0.06
    ल
    -0.06
    Manifest
    -0.06
    юдж
    -0.06
    -instance
    -0.06
    POSITIVE LOGITS
    along
    0.07
     six
    0.06
    丰
    0.06
     Riley
    0.06
     held
    0.06
    .dep
    0.06
    ResponseType
    0.06
    cant
    0.06
    복
    0.06
    (add
    0.06
    Activations Density 0.022%

    No Known Activations