© 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. 4-RES-SM
    11. 11768
    Prev
    Next
    INDEX
    Explanations

    terms related to invention and technological details

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    21
    +0.20
    1.2%
    456
    +0.12
    0.7%
    160
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.20
    0.04
    156
    +0.12
    0.03
    71
    +0.11
    0.03
    Negative Logits
    bie
    -1.75
    =%
    -1.70
    icals
    -1.62
     (@
    -1.57
    ^âĪĴ
    -1.52
    OOGLE
    -1.50
    bery
    -1.50
    arios
    -1.48
    rence
    -1.48
    ferroni
    -1.46
    POSITIVE LOGITS
    Patent
    2.14
    »¿
    2.02
     embodiments
    2.02
     invention
    1.95
    ©
    1.81
    ĵ
    1.75
    hereinafter
    1.75
    ŀ
    1.73
    Ī
    1.67
    This
    1.65
    Activations Density 0.034%

    No Known Activations