© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 1-RES-SM
    11. 26103
    Prev
    Next
    INDEX
    Explanations

    sequential numerical patterns or data series

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    410
    +0.12
    0.6%
    194
    +0.12
    0.6%
    511
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    325
    +0.12
    0.02
    273
    +0.12
    0.01
    100
    +0.12
    0.02
    Negative Logits
    »
    -2.13
    ļ
    -2.11
    »¿
    -2.01
    ĥ
    -1.95
    ĨĴ
    -1.86
    º
    -1.86
    ħ
    -1.84
    ł
    -1.79
    Ļª
    -1.78
    Ĥ¬
    -1.75
    POSITIVE LOGITS
    bian
    1.55
     plurality
    1.51
    forth
    1.49
      ↵
    1.49
    inator
    1.49
    iana
    1.47
    ppo
    1.46
    ierung
    1.45
    general
    1.43
    gerald
    1.43
    Activations Density 0.072%

    No Known Activations