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

    Neuronpedia

    Jacobian LensNEW
    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. 12714
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and formatting indicators within a text

    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
    21
    +0.32
    2.0%
    78
    +0.12
    0.8%
    101
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.32
    0.06
    78
    +0.12
    0.01
    397
    +0.10
    0.06
    Negative Logits
    idem
    -1.47
    lessly
    -1.43
     summation
    -1.32
     regard
    -1.29
     restraint
    -1.25
    itarian
    -1.23
    leys
    -1.23
    onna
    -1.21
    etin
    -1.20
    eks
    -1.20
    POSITIVE LOGITS
    ;
    2.96
    ;}
    2.53
    ;{\
    2.51
    ';
    2.39
    *;
    2.36
    ;"
    2.18
    .;
    2.16
    ;\
    2.15
    ";
    2.09
    ;,
    2.08
    Activations Density 0.927%

    No Known Activations