© 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. 100175
    Prev
    Next
    INDEX
    Explanations

    People

    np_max-act · gemini-2.0-flash

    references to people, particularly when it involves harm or danger.

    oai_token-act-pair · gpt-4oTriggered by @tcai

    This neuron responds to discussion of harm and safety, especially mentions of individuals, property, and well-being.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    requests for dangerous or illegal activities, particularly related to bomb-making.

    oai_token-act-pair · gpt-4o-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
     ERROR
    -0.07
    _schedule
    -0.07
    gif
    -0.07
    auction
    -0.07
     Selector
    -0.06
    rases
    -0.06
     deselect
    -0.06
    HSV
    -0.06
     Cellular
    -0.06
     ris
    -0.06
    POSITIVE LOGITS
    .assertIn
    0.07
    ันด
    0.07
     contexts
    0.07
    amel
    0.06
    MAR
    0.06
    (features
    0.06
    _pd
    0.06
    abad
    0.06
     plut
    0.06
     exclusively
    0.06
    Activations Density 0.034%

    No Known Activations