© 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. 27613
    Prev
    Next
    INDEX
    Explanations

    words related to effects and the concept of service

    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
    156
    +0.54
    4.0%
    111
    +0.47
    3.5%
    396
    +0.08
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    156
    +0.54
    0.23
    4
    +0.47
    0.23
    111
    +0.08
    0.21
    Negative Logits
    ´
    -1.78
    Ĵ
    -1.72
    Ľ
    -1.70
    ĵ
    -1.64
    ij
    -1.61
    ®
    -1.55
    <|padding|>
    -1.49
    ↵  ³³³
    -1.49
    -1.49
    ↵
    -1.49
    POSITIVE LOGITS
     aboard
    1.44
    aden
    1.34
     depending
    1.32
    unde
    1.29
    omyces
    1.20
    aline
    1.16
    uous
    1.15
    orems
    1.14
    ingly
    1.14
     {}
    1.14
    Activations Density 0.998%

    No Known Activations