© 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. 1-RES-SM
    11. 27221
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate relationships or connections between concepts

    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
    270
    +0.15
    0.8%
    151
    +0.12
    0.7%
    419
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    270
    +0.15
    0.01
    151
    +0.12
    0.01
    372
    +0.12
    0.01
    Negative Logits
    ļ
    -2.42
    Ĥ¬
    -2.37
    į
    -2.29
    ¶
    -2.28
    ®
    -2.25
    ¹
    -2.23
    ´
    -2.19
    Ģ
    -2.17
    ³
    -2.11
    ¨
    -2.08
    POSITIVE LOGITS
     advantages
    1.51
     disadvantages
    1.51
     superiority
    1.46
     favor
    1.45
     regards
    1.44
     men
    1.43
     wear
    1.43
    pecially
    1.39
     boys
    1.37
     sectional
    1.34
    Activations Density 0.034%

    No Known Activations