© 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. 13121
    Prev
    Next
    INDEX
    Explanations

    words and phrases that express a degree of abundance or supply

    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
    307
    +0.15
    0.9%
    450
    +0.12
    0.7%
    231
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    399
    +0.15
    0.02
    307
    +0.12
    0.03
    231
    +0.11
    0.03
    Negative Logits
    unnumbered
    -1.77
    ucker
    -1.70
    onym
    -1.62
    stitute
    -1.50
    itate
    -1.40
    ricks
    -1.40
    omething
    -1.39
    inher
    -1.38
    owl
    -1.37
    wers
    -1.37
    POSITIVE LOGITS
    ttes
    2.33
    ful
    2.16
    fully
    1.82
    ption
    1.79
    ncia
    1.78
    versible
    1.77
    tte
    1.74
    cean
    1.70
    gard
    1.70
    nel
    1.67
    Activations Density 0.317%

    No Known Activations