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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 55705
    Prev
    Next
    INDEX
    Explanations

    so

    np_max-act · gemini-2.0-flash

    This neuron activates specifically on the token “so,” identifying the discourse marker “so” in the text.

    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
    Bit
    -0.07
     Checked
    -0.06
     Symbol
    -0.06
    덕
    -0.06
    _supp
    -0.06
     starvation
    -0.06
     children
    -0.06
    .Padding
    -0.06
     Unused
    -0.06
    電視
    -0.06
    POSITIVE LOGITS
    ******↵↵
    0.07
    ?",
    0.06
    ,},↵
    0.06
     профилакти
    0.06
    so
    0.06
     =>'
    0.06
    FXML
    0.06
    ;(
    0.06
    .allow
    0.05
     },
    ↵
    ↵
    0.05
    Activations Density 0.035%

    No Known Activations