© 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
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 90706
    Prev
    Next
    INDEX
    Explanations

    Decepticon

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of Transformers faction names (especially “Decepticon,” and to a lesser extent “Autobot”).

    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
    یات
    -0.06
     dara
    -0.06
     пері
    -0.06
     stdout
    -0.06
    }}">{{$
    -0.06
    .paginator
    -0.06
    罗
    -0.06
    Ru
    -0.06
    ственное
    -0.06
     elapsed
    -0.06
    POSITIVE LOGITS
     Cycl
    0.07
     WRONG
    0.07
    part
    0.06
    esper
    0.06
     joked
    0.06
    ersh
    0.06
    REQUIRED
    0.06
    opc
    0.06
    -area
    0.06
    ushed
    0.06
    Activations Density 0.002%

    No Known Activations