© 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. 128522
    Prev
    Next
    INDEX
    Explanations

    is

    np_max-act · gemini-2.0-flash

    instructions or guidelines related to structured procedures or plans.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron fires on the initial filler tokens at the start of the assistant’s replies (e.g. “Here,” “is,” “a”), effectively spotting the “Here is a …” lead-in.

    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
    261
    -0.07
     NaN
    -0.06
    Checksum
    -0.06
     stimulation
    -0.06
     RectTransform
    -0.06
     bezier
    -0.06
     IEnumerable
    -0.06
    Gal
    -0.06
     Raw
    -0.06
    .ModelForm
    -0.06
    POSITIVE LOGITS
    送料
    0.08
    enting
    0.07
    _weak
    0.07
    ochond
    0.07
    oretical
    0.07
    _BAND
    0.06
    osemite
    0.06
     Provided
    0.06
    -colored
    0.06
    μέ
    0.06
    Activations Density 0.059%

    No Known Activations