© 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. 14036
    Prev
    Next
    INDEX
    Explanations

    components and configurations related to software packages

    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
    17
    +0.17
    1.0%
    431
    +0.12
    0.6%
    161
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    431
    +0.17
    0.05
    379
    +0.12
    0.08
    17
    +0.11
    0.01
    Negative Logits
     '</
    -1.87
    .[]{
    -1.80
    rapeut
    -1.78
     "</
    -1.73
    .]{}
    -1.68
     faint
    -1.64
    duly
    -1.58
     """
    -1.56
    $^{-
    -1.51
    ously
    -1.50
    POSITIVE LOGITS
    ĭ
    2.48
    ±
    2.25
    Ģ
    2.19
    ª
    2.17
    Ļª
    2.15
    ¢
    1.99
    ¶
    1.97
    ¦
    1.96
    ĨĴ
    1.94
    ³
    1.94
    Activations Density 0.548%

    No Known Activations