© 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
    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. 43451
    Prev
    Next
    INDEX
    Explanations

    positive adjectives

    np_max-act · gemini-2.0-flash

    This neuron activates on descriptive, evaluative adjectives (e.g. “elegant,” “engaging,” “immersive,” “melodic,” “dynamic”) that convey positive qualities or style.

    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
     acted
    -0.06
    Todd
    -0.06
     Addiction
    -0.06
    Videos
    -0.06
    膜
    -0.06
     Replace
    -0.06
    !/
    -0.06
    工業
    -0.06
     Defender
    -0.06
    €€
    -0.06
    POSITIVE LOGITS
     –↵
    0.07
     interacts
    0.06
     -↵↵
    0.06
    筆
    0.06
     ++;↵
    0.06
     υπ
    0.06
     lm
    0.06
     prima
    0.06
    ».↵
    0.06
    —↵↵
    0.06
    Activations Density 0.102%

    No Known Activations