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. 130370
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron detects hedging or qualifier phrases expressing tendency (e.g. “tends to …,” “rather …”).

    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
    中学
    -0.07
    fontsize
    -0.07
     Julie
    -0.06
    elen
    -0.06
    、三
    -0.06
     ACCOUNT
    -0.06
    ARB
    -0.06
    ニュ
    -0.06
     Proceed
    -0.06
     Seven
    -0.06
    POSITIVE LOGITS
     hoc
    0.07
    IS
    0.06
     [`
    0.06
     x
    0.06
    IAM
    0.06
    Updates
    0.06
     =>
    0.06
     types
    0.06
     skewed
    0.06
     trọng
    0.06
    Activations Density 0.015%

    No Known Activations