© 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. 30825
    Prev
    Next
    INDEX
    Explanations

    concepts related to legacy and historical impact

    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
    56
    +0.18
    1.1%
    450
    +0.15
    0.9%
    249
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    56
    +0.18
    -0.01
    100
    +0.15
    0.09
    249
    +0.11
    0.05
    Negative Logits
    ĨĴ
    -2.08
     yourself
    -2.08
    º
    -1.88
    Īĺ
    -1.84
    Ŀ
    -1.73
    IJ
    -1.72
    Ń
    -1.67
    ĺ
    -1.58
     yourselves
    -1.58
    Ĵ
    -1.57
    POSITIVE LOGITS
    opan
    1.83
     riches
    1.70
    itious
    1.57
    "}](#
    1.55
     credentials
    1.53
    velt
    1.52
    bsite
    1.49
     tales
    1.46
    áĢ¬
    1.45
     tale
    1.43
    Activations Density 1.626%

    No Known Activations