© 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. 1-RES-SM
    11. 23806
    Prev
    Next
    INDEX
    Explanations

    references to affiliations and partnerships

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.18
    1.0%
    367
    +0.15
    0.8%
    119
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    367
    +0.18
    0.03
    29
    +0.15
    0.01
    288
    +0.14
    0.02
    Negative Logits
    º
    -2.09
    ĨĴ
    -1.92
    ´
    -1.53
    ®
    -1.49
    ı
    -1.49
    Ĵ
    -1.46
    ¥
    -1.45
    ĺ
    -1.43
    Ķ
    -1.41
     transition
    -1.41
    POSITIVE LOGITS
     affiliations
    1.69
    nex
    1.62
     establishments
    1.60
    ially
    1.60
    het
    1.53
     organizations
    1.53
     oath
    1.51
    liament
    1.49
     organization
    1.47
     Institution
    1.46
    Activations Density 0.282%

    No Known Activations