© 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. 8577
    Prev
    Next
    INDEX
    Explanations

    instances of the word "the"

    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
    +0.29
    1.7%
    291
    +0.13
    0.8%
    368
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    291
    +0.29
    0.06
    145
    +0.13
    0.05
    10
    +0.13
    0.06
    Negative Logits
    ª
    -2.18
    º
    -2.17
    »
    -2.16
    ģ
    -2.13
    į
    -2.09
    Ģ
    -2.00
    ij
    -1.93
    ©
    -1.91
    ¹
    -1.89
    ¸
    -1.89
    POSITIVE LOGITS
     sake
    3.62
     purposes
    2.65
     entirety
    2.51
     foreseeable
    2.43
     rest
    2.10
     purpose
    2.05
     convenience
    2.02
     duration
    2.00
     same
    1.95
     better
    1.92
    Activations Density 0.403%

    No Known Activations