© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 32349
    Prev
    Next
    INDEX
    Explanations

    references to familial relationships and personal connections

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    56
    +0.23
    1.3%
    186
    +0.17
    1.0%
    423
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    56
    +0.23
    -0.00
    282
    +0.17
    0.10
    423
    +0.13
    0.14
    Negative Logits
     offline
    -1.80
    imits
    -1.63
    extend
    -1.58
    ipeline
    -1.57
    ITION
    -1.57
     quantification
    -1.56
    IAL
    -1.56
     Quantification
    -1.54
    >(
    -1.53
    ¯
    -1.53
    POSITIVE LOGITS
     baby
    2.29
     caregiver
    2.27
     mother
    2.17
     married
    2.14
     girl
    2.09
     daughter
    2.08
     twins
    2.06
     sick
    2.06
     babies
    2.04
     relatives
    1.99
    Activations Density 5.137%

    No Known Activations