© 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. 9716
    Prev
    Next
    INDEX
    Explanations

    references to a concluding or final stage in various contexts

    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
    100
    +0.17
    1.0%
    365
    +0.13
    0.7%
    489
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    100
    +0.17
    0.03
    206
    +0.13
    0.01
    476
    +0.11
    0.03
    Negative Logits
    chitz
    -2.21
    lav
    -1.73
    mania
    -1.71
    fulness
    -1.67
    roz
    -1.60
    ãĥ¥
    -1.58
    ÏħÏĦ
    -1.52
    openh
    -1.52
    aland
    -1.51
    urop
    -1.50
    POSITIVE LOGITS
     tally
    2.09
    ized
    1.87
    ists
    1.85
    isations
    1.81
    izing
    1.81
    ised
    1.78
     submission
    1.75
     decree
    1.74
    ising
    1.73
    izers
    1.70
    Activations Density 0.110%

    No Known Activations