© 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. 12991
    Prev
    Next
    INDEX
    Explanations

    references to numeric values and statistical data

    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
    156
    +0.34
    2.2%
    111
    +0.28
    1.8%
    271
    +0.20
    1.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    266
    +0.34
    0.06
    271
    +0.28
    0.01
    148
    +0.20
    0.02
    Negative Logits
    pson
    -1.47
    ).)
    -1.43
     aid
    -1.40
     counterpart
    -1.39
     structures
    -1.38
    iaz
    -1.35
    bound
    -1.35
    bol
    -1.31
    ized
    -1.30
    pel
    -1.29
    POSITIVE LOGITS
    Į
    4.00
    ¼
    3.83
    ½
    3.70
    Ī
    3.70
    »¿
    3.49
    ĻĤ
    3.45
    »
    3.33
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    3.25
    į
    3.23
    Ĥ
    3.22
    Activations Density 0.476%

    No Known Activations