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. 130481
    Prev
    Next
    INDEX
    Explanations

    expressing personal opinions

    np_max-act · gemini-2.0-flash

    The neuron responds to subjective emphasis in first‐person statements, firing on pronouns and intensifiers like “me,” “bit,” “lot,” and “really.”

    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
     driveway
    -0.07
    �
    -0.07
     të
    -0.07
    -heading
    -0.07
     Title
    -0.06
     gec
    -0.06
    ncy
    -0.06
    (",");↵
    -0.06
    -spin
    -0.06
     nowrap
    -0.06
    POSITIVE LOGITS
    Azure
    0.06
     WINDOWS
    0.06
    Closure
    0.06
    }while
    0.06
     místní
    0.06
    .Future
    0.06
     çok
    0.06
    Reaction
    0.06
    eguard
    0.06
    ARSER
    0.06
    Activations Density 0.277%

    No Known Activations