© 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 EvalsExportsinterp-engineNEWAPI 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. 31184
    Prev
    Next
    INDEX
    Explanations

    the verb "be" in various forms and usages

    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
    419
    +0.14
    0.8%
    263
    +0.13
    0.7%
    272
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    272
    +0.14
    0.06
    260
    +0.13
    0.05
    435
    +0.13
    0.05
    Negative Logits
    İ
    -3.85
    ī
    -3.80
    Ī
    -3.76
    ¥
    -3.74
    Ŀ
    -3.65
    ŀ
    -3.63
    Ĭ
    -3.56
    ł
    -3.54
    ¡
    -3.52
    Ľ
    -3.52
    POSITIVE LOGITS
    pez
    2.06
     anywhere
    1.87
    aucoup
    1.76
    zos
    1.76
    yon
    1.74
     particularly
    1.71
    ardless
    1.62
    quez
    1.62
    aded
    1.62
     interpreted
    1.59
    Activations Density 0.194%

    No Known Activations