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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 11862
    Prev
    Next
    INDEX
    Explanations

    programming patterns related to file management and data processing in Python

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    397
    +0.20
    1.1%
    78
    +0.14
    0.8%
    408
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    397
    +0.20
    0.12
    78
    +0.14
    0.01
    423
    +0.13
    0.11
    Negative Logits
    ";
    -2.11
    ”;
    -2.04
    ");
    -1.78
    .";
    -1.72
    !");
    -1.70
    ;
    -1.70
    ABILITY
    -1.69
     plaintiffs
    -1.69
     ;
    -1.62
    "));
    -1.61
    POSITIVE LOGITS
    umpy
    2.48
     numpy
    2.26
     """
    2.26
    ŀ
    2.26
    Ŀ
    2.24
     {'
    2.24
     np
    2.19
     kw
    2.18
    ."""
    2.17
    »
    2.16
    Activations Density 1.323%

    No Known Activations