© 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. 1-RES-SM
    11. 25601
    Prev
    Next
    INDEX
    Explanations

    complex mathematical expressions or calculations

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    398
    +0.13
    0.7%
    510
    +0.10
    0.6%
    453
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    398
    +0.13
    0.02
    499
    +0.10
    0.01
    496
    +0.10
    0.02
    Negative Logits
    ĻĤ
    -2.28
    ³
    -1.95
    ...](
    -1.63
    xymatrix
    -1.58
    °
    -1.53
    į
    -1.45
    Ĺ
    -1.37
    -1.35
      ↵
    -1.35
    -1.35
    POSITIVE LOGITS
     denomin
    1.62
     lett
    1.52
    iar
    1.46
    apiro
    1.41
    ott
    1.37
     dividend
    1.35
     PBX
    1.33
    ository
    1.32
     drafts
    1.32
    velop
    1.31
    Activations Density 0.075%

    No Known Activations