© 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. 41657
    Prev
    Next
    INDEX
    Explanations

    frenzied excitement

    np_max-act · gemini-2.0-flash

    This neuron activates on descriptive modifiers—especially adverbs and intensity adjectives (like “tightly” or “frenzied”) that describe how actions are performed.

    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
     cube
    -0.07
    (com
    -0.06
    Dll
    -0.06
    "bytes
    -0.06
    }while
    -0.06
     NotImplemented
    -0.06
     ultra
    -0.06
    .Mesh
    -0.06
    -compatible
    -0.06
    (cps
    -0.06
    POSITIVE LOGITS
     frantic
    0.07
     Lawn
    0.07
     attempts
    0.07
     furious
    0.07
     lawn
    0.07
     режим
    0.06
     earnest
    0.06
     escort
    0.06
    慌
    0.06
    'Connor
    0.06
    Activations Density 0.014%

    No Known Activations