© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 43337
    Prev
    Next
    INDEX
    Explanations

    general english

    np_max-act · gemini-2.0-flash

    This neuron detects words involved in negating or refuting causal claims, especially disclaiming intentional design or action by agents (e.g., “not,” “because,” “intentional,” “by aliens”).

    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
    enc
    -0.06
    MATCH
    -0.06
     teşkil
    -0.06
    MaxLength
    -0.06
    γου
    -0.06
     แ
    -0.06
    옥
    -0.06
    وئ
    -0.06
     websites
    -0.06
     chir
    -0.06
    POSITIVE LOGITS
    .DataFrame
    0.07
     russian
    0.06
    resultado
    0.06
     DAMAGES
    0.06
    -ver
    0.06
    LocalizedMessage
    0.06
     Diary
    0.06
    ---------------↵
    0.06
    ..↵↵
    0.06
    (~
    0.06
    Activations Density 0.114%

    No Known Activations