© 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. 24359
    Prev
    Next
    INDEX
    Explanations

    references to specific years, particularly 2017 and its subsequent trends

    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
    457
    +0.17
    1.0%
    51
    +0.16
    0.9%
    246
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    51
    +0.17
    0.04
    308
    +0.16
    0.04
    176
    +0.13
    0.04
    Negative Logits
    Ł
    -1.71
    Ļª
    -1.62
    º
    -1.59
    ij
    -1.58
     control
    -1.57
     verb
    -1.54
     zone
    -1.48
    ŀ
    -1.46
     jurisdiction
    -1.46
    ©
    -1.43
    POSITIVE LOGITS
    uscript
    1.60
     costs
    1.60
     expenditure
    1.58
     achieved
    1.53
     expenditures
    1.51
    å¹´
    1.50
    latest
    1.48
    views
    1.42
     recipients
    1.42
    #{$
    1.41
    Activations Density 0.083%

    No Known Activations