© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 59335
    Prev
    Next
    INDEX
    Explanations

    Text structure

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on common English function words and punctuation (e.g. “is,” “to,” “they,” commas), effectively marking high‐frequency, low‐content tokens.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     Animator
    -0.06
     positional
    -0.06
    oppel
    -0.06
     derivative
    -0.06
     adolescente
    -0.05
    :::::::::::
    -0.05
    parseInt
    -0.05
    ptune
    -0.05
    .Collapsed
    -0.05
    LI
    -0.05
    POSITIVE LOGITS
    洲
    0.07
    یا
    0.07
    uya
    0.07
    Invoice
    0.07
    υχ
    0.07
     yö
    0.07
     SignIn
    0.07
    .todo
    0.06
     Queensland
    0.06
    monic
    0.06
    Activations Density 0.062%

    No Known Activations