© 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 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. 10809
    Prev
    Next
    INDEX
    Explanations

    instances of actions related to publishing or posting content

    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
    156
    +0.21
    1.2%
    266
    +0.11
    0.6%
    376
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    266
    +0.21
    0.06
    304
    +0.11
    0.05
    339
    +0.10
    0.03
    Negative Logits
    Ĥ
    -1.94
    °
    -1.93
    ¢
    -1.82
    ¶
    -1.70
    ĥ
    -1.65
    Īĺ
    -1.63
    º
    -1.60
    ľ
    -1.59
    Ĩ
    -1.53
    ·
    -1.47
    POSITIVE LOGITS
     by
    2.62
     By
    2.08
     Parenthood
    2.06
     from
    1.86
    lasses
    1.75
    izione
    1.71
     pursuant
    1.69
    river
    1.67
     via
    1.65
    itors
    1.64
    Activations Density 0.816%

    No Known Activations