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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 1050
    Prev
    Next
    INDEX
    Explanations

    Questions and prompts

    np_max-act · gemini-2.0-flash

    This neuron fires on the numeric length‐specifier tokens in user prompts (e.g. the “200,” “100,” or “500” in “200-word,” “100-word,” etc.).

    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
    .points
    -0.07
     Doing
    -0.07
    ,C
    -0.06
     dots
    -0.06
     méd
    -0.06
    .good
    -0.06
    _Manager
    -0.06
    рел
    -0.06
    zero
    -0.06
     spot
    -0.06
    POSITIVE LOGITS
    ictureBox
    0.06
    cwd
    0.06
    РСР
    0.06
    降
    0.06
    ceiver
    0.06
     devoid
    0.06
     تمام
    0.06
     dotenv
    0.06
    ув
    0.06
     оказ
    0.06
    Activations Density 0.015%

    No Known Activations