© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 30518
    Prev
    Next
    INDEX
    Explanations

    /

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on punctuation markers (parentheses and slashes) that introduce alternate names, abbreviations, or synonyms.

    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
    денти
    -0.07
     codec
    -0.07
    -forward
    -0.06
     practitioners
    -0.06
    ,x
    -0.06
     gelecek
    -0.06
    ethical
    -0.06
    ึกษา
    -0.06
    ео
    -0.06
     EnumerableStream
    -0.06
    POSITIVE LOGITS
    	point
    0.07
     muy
    0.07
     bang
    0.07
    illation
    0.07
     branching
    0.06
    	IN
    0.06
    grab
    0.06
    .command
    0.06
     مرد
    0.06
     Thanksgiving
    0.06
    Activations Density 0.046%

    No Known Activations