© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 30832
    Prev
    Next
    INDEX
    Explanations

    not

    np_max-act · gemini-2.0-flash

    The neuron spikes on negation cues—especially the word “not” (and related exclusivity markers like “only”)—flagging negative statements.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    text that discusses failures and criticisms in sports performance and management.

    oai_token-act-pair · gpt-4o-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
     Drama
    -0.07
    622
    -0.06
    /IP
    -0.06
    (cf
    -0.06
     obtain
    -0.06
     Within
    -0.06
    601
    -0.06
     supply
    -0.06
     scape
    -0.06
     <<
    -0.06
    POSITIVE LOGITS
     not
    0.08
    ưa
    0.07
    _FORCE
    0.07
    .setStyleSheet
    0.07
     NOT
    0.07
    هل
    0.07
     توص
    0.07
    /not
    0.07
     mostly
    0.06
    inking
    0.06
    Activations Density 0.025%

    No Known Activations