© 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. 29996
    Prev
    Next
    INDEX
    Explanations

    references to specific details or criteria

    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
    362
    +0.12
    0.7%
    438
    +0.12
    0.6%
    327
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    424
    +0.12
    0.02
    179
    +0.12
    0.02
    438
    +0.11
    0.02
    Negative Logits
    ulic
    -2.12
    oming
    -1.69
    lementary
    -1.64
    heimer
    -1.58
    oard
    -1.55
    rein
    -1.52
    emon
    -1.50
    inth
    -1.48
    uler
    -1.47
    smallcaps
    -1.47
    POSITIVE LOGITS
    Ł
    2.34
    ities
    2.31
    Ń
    2.21
    ¥
    2.20
    Ŀ
    2.12
    £
    2.12
    Īĺ
    2.04
    ences
    1.99
     aims
    1.97
    ¨
    1.94
    Activations Density 0.073%

    No Known Activations