© 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. 13285
    Prev
    Next
    INDEX
    Explanations

    references to processes that involve using or applying something, particularly the term "to" indicating intent or action

    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
    431
    +0.13
    0.7%
    322
    +0.12
    0.7%
    344
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    434
    +0.13
    0.05
    322
    +0.12
    0.06
    393
    +0.11
    0.02
    Negative Logits
    »¿
    -2.58
    ¼
    -2.37
    ¡
    -1.98
    ĻĤ
    -1.89
    ÏīÏĤ
    -1.85
    Ī
    -1.84
    ®
    -1.83
    Ļ
    -1.82
    ª
    -1.79
    ĩ
    -1.75
    POSITIVE LOGITS
     treat
    2.33
     describe
    2.33
     prepare
    2.27
     simulate
    2.27
     recreate
    2.24
     calculate
    2.21
     capture
    2.21
     manage
    2.20
     create
    2.20
     assess
    2.20
    Activations Density 0.258%

    No Known Activations