© 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. 10742
    Prev
    Next
    INDEX
    Explanations

    numerical data and formats related to various subjects

    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
    +0.23
    1.4%
    186
    +0.17
    1.1%
    320
    +0.17
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    419
    +0.23
    0.10
    271
    +0.17
    -0.05
    27
    +0.17
    0.10
    Negative Logits
    Ļª
    -1.98
    ĨĴ
    -1.92
    ij
    -1.90
    Ĥ¬
    -1.70
    ¬
    -1.50
    ĻĤ
    -1.42
    ¾
    -1.41
    IJ
    -1.40
     Dictionary
    -1.39
    ictionary
    -1.35
    POSITIVE LOGITS
    иб
    1.74
    еÑģк
    1.45
      ↵
    1.37
     whom
    1.36
    keit
    1.30
    Willi
    1.30
     hell
    1.29
    Ïģε
    1.28
     whose
    1.27
    ski
    1.27
    Activations Density 4.761%

    No Known Activations