© 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. 13365
    Prev
    Next
    INDEX
    Explanations

    the definite article "the" in various contexts

    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
    +0.28
    1.6%
    422
    +0.12
    0.7%
    478
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    98
    +0.28
    0.01
    291
    +0.12
    0.01
    64
    +0.12
    0.02
    Negative Logits
    nos
    -1.88
    erty
    -1.78
    cedent
    -1.75
    ally
    -1.70
    rolet
    -1.65
    ching
    -1.64
    pot
    -1.57
    ?).
    -1.57
    rying
    -1.56
    vester
    -1.54
    POSITIVE LOGITS
    Ī
    3.43
    ¬
    3.42
    ĸ´
    3.35
    IJ
    3.31
    ¼
    3.30
    ĥ
    3.29
    Ĥ
    3.17
    ¸
    3.13
    Ĩ
    3.12
    «
    3.09
    Activations Density 0.043%

    No Known Activations