© 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. 27567
    Prev
    Next
    INDEX
    Explanations

    programming-related syntax and structures, particularly imports and exports in code

    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
    93
    +0.15
    0.8%
    42
    +0.12
    0.6%
    491
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    131
    +0.15
    0.03
    184
    +0.12
    0.02
    43
    +0.11
    0.02
    Negative Logits
    enron
    -1.68
    iNdEx
    -1.67
    mouth
    -1.66
    zia
    -1.59
    goto
    -1.55
    hop
    -1.49
    rceil
    -1.45
    overs
    -1.43
    peg
    -1.43
    interrupt
    -1.42
    POSITIVE LOGITS
    ħ
    1.93
    ¯
    1.70
    ¦
    1.69
     panels
    1.68
     Caption
    1.67
    ľ
    1.58
    ¸
    1.56
     Commerce
    1.53
    ¶
    1.53
    ·¸
    1.46
    Activations Density 0.103%

    No Known Activations