© 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. 2-RES-SM
    11. 28107
    Prev
    Next
    INDEX
    Explanations

    patterns of numerical values, particularly those formatted similarly to currency or counting systems

    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
    98
    +0.13
    0.8%
    287
    +0.13
    0.8%
    136
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    136
    +0.13
    -0.00
    167
    +0.13
    0.08
    287
    +0.12
    0.13
    Negative Logits
    Ŀ
    -1.94
    ¹
    -1.88
    °
    -1.88
    ŀ
    -1.85
    ı
    -1.85
    ¬
    -1.71
    ©
    -1.71
    ¦
    -1.71
    IJ
    -1.71
    Ĩ
    -1.66
    POSITIVE LOGITS
    arynge
    1.89
    GREEK
    1.41
    eeper
    1.37
    onant
    1.36
    astric
    1.35
    veolar
    1.35
    ÈĻ
    1.35
    ellees
    1.34
    ellants
    1.33
    odon
    1.32
    Activations Density 2.326%

    No Known Activations