© 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. 7531
    Prev
    Next
    INDEX
    Explanations

    classical music compositions

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on numeric metadata and descriptors in compositional write-ups—things like dates, durations, opus numbers, keys, and other numerical details in musical work descriptions.

    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
    LEMENT
    -0.06
     rotational
    -0.06
     paid
    -0.06
    }px
    -0.05
     retorna
    -0.05
    continuous
    -0.05
    -instagram
    -0.05
     Fac
    -0.05
    chet
    -0.05
    аніз
    -0.05
    POSITIVE LOGITS
     chứ
    0.07
    ότητα
    0.07
     computer
    0.07
     rifles
    0.07
    hec
    0.06
    대학교
    0.06
    pun
    0.06
     мон
    0.06
    ğiniz
    0.06
     race
    0.06
    Activations Density 0.016%

    No Known Activations