© 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. 32415
    Prev
    Next
    INDEX
    Explanations

    references to renewable energy sources and their benefits

    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
    404
    +0.12
    0.7%
    266
    +0.11
    0.6%
    274
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    404
    +0.12
    0.07
    88
    +0.11
    0.07
    238
    +0.11
    0.05
    Negative Logits
    ^](#
    -1.75
    pect
    -1.62
     briefs
    -1.58
     wines
    -1.53
     charts
    -1.50
     notes
    -1.49
     figures
    -1.46
     horiz
    -1.44
     beers
    -1.38
     Wine
    -1.35
    POSITIVE LOGITS
    ľĵ
    2.14
    Į
    1.96
    ĸ
    1.83
    endo
    1.55
    adding
    1.50
    ·
    1.47
    ij
    1.46
    casting
    1.46
    pgen
    1.46
    saving
    1.43
    Activations Density 0.499%

    No Known Activations