© 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. 4-RES-SM
    11. 8756
    Prev
    Next
    INDEX
    Explanations

    numerical values and their context regarding financial or statistical information

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.14
    0.8%
    186
    +0.11
    0.6%
    227
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    30
    +0.14
    0.01
    333
    +0.11
    0.05
    131
    +0.11
    0.05
    Negative Logits
     typedef
    -1.52
    asan
    -1.43
     resignation
    -1.42
     Ãİ
    -1.41
    ÈĻ
    -1.39
    quire
    -1.39
     implied
    -1.39
    unnumbered
    -1.38
     \|_{
    -1.38
     Bour
    -1.37
    POSITIVE LOGITS
     500
    3.03
     1000
    3.02
     200
    2.94
     600
    2.93
     400
    2.92
     300
    2.88
     1500
    2.82
     5000
    2.80
     800
    2.77
     100
    2.77
    Activations Density 1.376%

    No Known Activations