© 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. 27597
    Prev
    Next
    INDEX
    Explanations

    vowel combinations, particularly those with 'e' and 'a'

    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
    241
    +0.15
    0.9%
    144
    +0.12
    0.7%
    92
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.15
    0.03
    271
    +0.12
    -0.00
    144
    +0.12
    0.02
    Negative Logits
    ĨĴ
    -2.06
    stage
    -1.72
    ĵ
    -1.65
    ©
    -1.54
    lining
    -1.51
    aturally
    -1.50
    itary
    -1.49
    translation
    -1.43
     moderation
    -1.41
    ensibly
    -1.40
    POSITIVE LOGITS
    zing
    2.15
    ppo
    1.96
    zes
    1.96
    lectric
    1.92
    hler
    1.79
    cember
    1.79
    chter
    1.79
    vel
    1.77
    fficients
    1.74
    ings
    1.66
    Activations Density 0.222%

    No Known Activations