© 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. 109476
    Prev
    Next
    INDEX
    Explanations

    technical texts

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on hyphenated or multi-part technical modifiers (e.g. multi-stage, low-velocity, linear-ly, axially-and-rotational, parallel-plate, plano-convex).

    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
    іть
    -0.07
     หาก
    -0.07
     diğer
    -0.06
    ReadOnly
    -0.06
    _RULE
    -0.06
     vytvá
    -0.06
    itic
    -0.06
    _hour
    -0.06
     Lý
    -0.06
    processed
    -0.06
    POSITIVE LOGITS
    POSITORY
    0.06
    (classes
    0.06
     zombies
    0.06
    abcdefgh
    0.06
    (Address
    0.06
    '+
    0.06
    (script
    0.06
    Svc
    0.06
    .repository
    0.06
    食べ
    0.06
    Activations Density 0.186%

    No Known Activations