© 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. 68527
    Prev
    Next
    INDEX
    Explanations

    Code and math

    np_max-act · gemini-2.0-flash

    This neuron activates on the phrasing of the consistency‐check question—especially tokens like “Is,” “the,” and “respect”—marking where the text asks whether the summary is factually consistent.

    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
     admit
    -0.07
     youth
    -0.07
    clipse
    -0.06
    -wide
    -0.06
    يرة
    -0.06
    vůli
    -0.06
    academic
    -0.06
     �
    -0.06
     Před
    -0.06
     main
    -0.06
    POSITIVE LOGITS
     Record
    0.08
     dab
    0.07
    舒
    0.07
     xây
    0.07
     suche
    0.06
    .reflect
    0.06
     overseeing
    0.06
     کتاب
    0.06
     Não
    0.06
     tad
    0.06
    Activations Density 0.003%

    No Known Activations