© 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. 26274
    Prev
    Next
    INDEX
    Explanations

    references to data structures and programming concepts

    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
    78
    +0.13
    0.8%
    187
    +0.13
    0.7%
    320
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    185
    +0.13
    0.07
    431
    +0.13
    0.09
    187
    +0.12
    0.08
    Negative Logits
    '));
    -1.62
    Willi
    -1.55
    oli
    -1.51
    astric
    -1.48
    inian
    -1.47
    ella
    -1.45
    etta
    -1.44
    iest
    -1.44
    throw
    -1.43
    lique
    -1.43
    POSITIVE LOGITS
    ¥
    1.73
     degrees
    1.64
     participant
    1.53
    ments
    1.50
    ¾
    1.42
    Ŀ
    1.41
    ķ
    1.35
     categorical
    1.34
     Awards
    1.33
     nom
    1.30
    Activations Density 4.349%

    No Known Activations