© 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. 12344
    Prev
    Next
    INDEX
    Explanations

    references to error handling and file processing

    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
    156
    +0.18
    1.0%
    78
    +0.16
    0.9%
    36
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    9
    +0.18
    0.04
    456
    +0.16
    0.06
    397
    +0.15
    0.04
    Negative Logits
     carrier
    -1.40
     away
    -1.36
    arians
    -1.34
     oneself
    -1.34
     suitability
    -1.33
     Bachelor
    -1.32
     reversal
    -1.31
     idea
    -1.31
     derivatives
    -1.28
    amycin
    -1.27
    POSITIVE LOGITS
    ī
    3.82
    ²
    3.77
    ¤
    3.72
    ĨĴ
    3.53
    ¬
    3.38
    Ĵ
    3.33
    Ĥ¬
    3.31
    Īĺ
    3.30
    ģ
    3.28
    Ľ
    3.23
    Activations Density 0.366%

    No Known Activations