© 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. 14025
    Prev
    Next
    INDEX
    Explanations

    words related to chemical processes or reactions

    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
    369
    +0.14
    0.8%
    377
    +0.12
    0.6%
    209
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    209
    +0.14
    0.04
    377
    +0.12
    0.04
    125
    +0.12
    0.03
    Negative Logits
    ingale
    -1.59
     kidding
    -1.54
    ESULT
    -1.53
    born
    -1.47
    lace
    -1.41
    eren
    -1.41
    º
    -1.41
     whole
    -1.39
    er
    -1.38
    feit
    -1.37
    POSITIVE LOGITS
    ritz
    1.92
    ructose
    1.80
    ahren
    1.79
    inally
    1.79
    aux
    1.78
    erguson
    1.76
    licts
    1.73
    acet
    1.69
    requently
    1.68
    uses
    1.64
    Activations Density 0.103%

    No Known Activations