Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 16167
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    This neuron selectively activates on content-bearing subword tokens (longer semantic stems) rather than on common function words.

    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
    !!");↵
    -0.07
    )});↵
    -0.06
     mosquitoes
    -0.06
    ียม
    -0.06
    ,即
    -0.06
    ською
    -0.06
    '))->
    -0.06
     calculations
    -0.06
    ")).
    -0.06
    .");↵
    -0.06
    POSITIVE LOGITS
    Genre
    0.06
     Abdul
    0.06
    844
    0.06
     выраж
    0.06
    .bi
    0.06
    _argv
    0.06
     sociales
    0.06
    脸
    0.06
     TRAIN
    0.06
    /job
    0.06
    Activations Density 0.216%

    No Known Activations