© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 27557
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "I'm" or variations indicating personal statements or confessions

    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
    410
    +0.17
    0.9%
    171
    +0.11
    0.6%
    478
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    171
    +0.17
    0.05
    348
    +0.11
    0.06
    250
    +0.11
    0.06
    Negative Logits
     ordinary
    -1.72
     <-
    -1.64
    onymous
    -1.63
     ![](
    -1.50
    ictional
    -1.48
    former
    -1.38
    ('/
    -1.38
    oddsidemargin
    -1.38
    ()">
    -1.35
     charity
    -1.34
    POSITIVE LOGITS
    s
    1.90
    ÃĤ
    1.86
    keit
    1.66
    ¾
    1.64
    mie
    1.64
    MENT
    1.62
     regards
    1.59
    zed
    1.55
    emi
    1.52
    ŀ
    1.51
    Activations Density 0.170%

    No Known Activations