© 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. 7770
    Prev
    Next
    INDEX
    Explanations

    cost

    np_max-act · gemini-2.0-flash

    The neuron is looking for occurrences of the word “cost” (including compounds like “cost-effectiveness”) in the text.

    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
    ruz
    -0.07
    rawl
    -0.07
    ává
    -0.07
     печ
    -0.07
    věd
    -0.07
    22
    -0.07
     bubble
    -0.07
    ัม
    -0.07
     veh
    -0.07
     around
    -0.07
    POSITIVE LOGITS
     cost
    0.15
     Cost
    0.14
    Cost
    0.14
     costs
    0.14
    cost
    0.14
     COST
    0.13
     Costs
    0.12
     costing
    0.11
    _cost
    0.10
    .cost
    0.09
    Activations Density 0.030%

    No Known Activations