© 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. 26806
    Prev
    Next
    INDEX
    Explanations

    instances of specific formatting or structural elements in the text

    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
    111
    +0.78
    16.1%
    156
    +0.60
    12.4%
    71
    +0.05
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.78
    0.85
    156
    +0.60
    0.72
    71
    +0.05
    0.69
    Negative Logits
    ¿½
    -2.58
    ħ
    -2.53
    ľĵ
    -2.46
    Ľ
    -2.46
    ij
    -2.45
    ¿
    -2.43
    ®
    -2.42
    ¦
    -2.41
    Ģ
    -2.39
    ½
    -2.37
    POSITIVE LOGITS
     polÃŃ
    0.95
     fundament
    0.94
     invalid
    0.90
     aboard
    0.90
     inheritance
    0.88
     hype
    0.87
     opinions
    0.85
    être
    0.85
     concentration
    0.85
    orph
    0.84
    Activations Density 0.905%

    No Known Activations