© 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. 47922
    Prev
    Next
    INDEX
    Explanations

    had

    np_max-act · gemini-2.0-flash

    events related to governance and political negotiations.

    oai_token-act-pair · gpt-4o-mini
    Triggered by @xinyanhu8

    The neuron is primarily responding to verb forms—especially past-tense and modal/auxiliary verbs (e.g. “had,” “was,” “said,” “could,” “would”).

    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
     Exclusive
    -0.07
    950
    -0.07
    IFI
    -0.06
    скую
    -0.06
    pdf
    -0.06
     attributeName
    -0.06
     precipitation
    -0.06
    fic
    -0.06
    cookie
    -0.06
     instructions
    -0.06
    POSITIVE LOGITS
     leaned
    0.07
     sebuah
    0.06
    	ASSERT
    0.06
    Jackson
    0.06
     artış
    0.06
     peri
    0.06
    >'
↵
    0.06
     Ebony
    0.06
     compart
    0.06
    .isDebugEnabled
    0.06
    Activations Density 0.028%

    No Known Activations