© 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. 14163
    Prev
    Next
    INDEX
    Explanations

    references to potential risks or warnings in discussions about products or concepts

    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
    83
    +0.12
    0.7%
    39
    +0.11
    0.6%
    159
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    450
    +0.12
    0.03
    209
    +0.11
    0.04
    273
    +0.10
    0.03
    Negative Logits
    ľĵ
    -1.96
    Ĺ
    -1.92
    emat
    -1.85
    Ļ
    -1.84
    »¿
    -1.79
    à±į
    -1.75
    Ħ
    -1.73
    ĸ
    -1.70
    Ĩ
    -1.69
    ij
    -1.61
    POSITIVE LOGITS
    roup
    1.91
    ============
    1.56
    asone
    1.52
     Clause
    1.44
     subclass
    1.38
     sake
    1.35
    bbe
    1.34
     Sart
    1.32
    =========
    1.31
    laws
    1.30
    Activations Density 0.221%

    No Known Activations