© 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. 26431
    Prev
    Next
    INDEX
    Explanations

    elements formatted as data entries or parameters in a structured format

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    369
    +0.19
    1.1%
    436
    +0.17
    0.9%
    478
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    136
    +0.19
    -0.02
    332
    +0.17
    0.06
    436
    +0.13
    0.07
    Negative Logits
    º
    -2.00
    Ŀ
    -1.92
    »
    -1.92
    ģ
    -1.89
    igraphy
    -1.86
    ĸ
    -1.80
    ¨
    -1.75
    aper
    -1.71
    Ĺ
    -1.67
    ĸ´
    -1.67
    POSITIVE LOGITS
     Bund
    1.60
    denly
    1.45
     Poor
    1.41
    iliar
    1.37
     SPD
    1.34
     redirect
    1.32
     ":
    1.32
    uced
    1.30
     insistence
    1.30
    FU
    1.29
    Activations Density 1.711%

    No Known Activations