© 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. 32486
    Prev
    Next
    INDEX
    Explanations

    the indefinite articles and their frequency

    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
    98
    +0.16
    0.9%
    478
    +0.14
    0.8%
    436
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    137
    +0.16
    0.10
    123
    +0.14
    0.09
    321
    +0.13
    0.07
    Negative Logits
    enance
    -1.72
     alloc
    -1.65
    bels
    -1.64
     encounters
    -1.61
     prices
    -1.60
    walls
    -1.56
    iances
    -1.56
    lications
    -1.53
    ventory
    -1.48
    arms
    -1.46
    POSITIVE LOGITS
     lot
    1.91
     kind
    1.90
     priori
    1.83
     combination
    1.79
     mixture
    1.78
     flaw
    1.78
     bit
    1.75
    '?
    1.75
    plasia
    1.73
     purely
    1.71
    Activations Density 0.800%

    No Known Activations