© 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. 109009
    Prev
    Next
    INDEX
    Explanations

    пере

    np_max-act · gemini-2.0-flash

    The neuron detects Russian verbs formed with the prefix “пере-” (e.g. перескажи, переключиться, перемещается).

    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
     Doc
    -0.07
     Holy
    -0.07
     Stout
    -0.07
     councillor
    -0.07
     Folk
    -0.07
     lé
    -0.07
     districts
    -0.07
     duty
    -0.06
     party
    -0.06
     род
    -0.06
    POSITIVE LOGITS
    .twig
    0.07
    ifferent
    0.07
     reim
    0.07
    .Here
    0.07
     перен
    0.07
     permutations
    0.07
    eyh
    0.07
     parenthesis
    0.07
     перевір
    0.07
    .Net
    0.07
    Activations Density 0.012%

    No Known Activations