© 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. 32430
    Prev
    Next
    INDEX
    Explanations

    numerical values related to game scores and dates

    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
    342
    +0.13
    0.7%
    198
    +0.12
    0.7%
    64
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    342
    +0.13
    0.06
    477
    +0.12
    0.06
    427
    +0.11
    0.04
    Negative Logits
    vae
    -1.65
    ities
    -1.38
    uities
    -1.33
    ques
    -1.31
     pain
    -1.30
     Powers
    -1.26
     =&
    -1.23
     Von
    -1.23
    esis
    -1.23
    angs
    -1.23
    POSITIVE LOGITS
    room
    1.83
    ubuntu
    1.69
    hire
    1.57
    bed
    1.54
    erto
    1.52
    addy
    1.51
    tal
    1.49
    riterion
    1.44
    ked
    1.43
    sted
    1.42
    Activations Density 0.254%

    No Known Activations