© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 4-RES-SM
    11. 12301
    Prev
    Next
    INDEX
    Explanations

    mentions of the letter 'D'

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    369
    +0.13
    0.7%
    271
    +0.12
    0.7%
    109
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    369
    +0.13
    0.03
    109
    +0.12
    0.04
    296
    +0.11
    0.04
    Negative Logits
     oneself
    -1.52
    lets
    -1.49
     ourselves
    -1.45
    ptions
    -1.43
     advantage
    -1.40
    places
    -1.39
    bars
    -1.39
    selves
    -1.39
     always
    -1.37
    borough
    -1.35
    POSITIVE LOGITS
    Ļª
    3.82
    ®
    3.75
    ´
    3.54
    ĨĴ
    3.40
    Ļ
    3.38
    ¤
    3.19
    ¶
    3.17
    Ń
    3.17
    ľ
    3.14
    ³
    3.11
    Activations Density 0.087%

    No Known Activations