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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 11392
    Prev
    Next
    INDEX
    Explanations

    phrases related to measurement and evaluation in scientific studies

    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
    478
    +0.14
    0.8%
    85
    +0.13
    0.7%
    225
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    88
    +0.14
    0.03
    404
    +0.13
    0.03
    238
    +0.11
    0.03
    Negative Logits
    holder
    -1.54
     credits
    -1.46
    stown
    -1.42
     disputes
    -1.36
     Soccer
    -1.34
     Tobacco
    -1.34
    verb
    -1.33
    blance
    -1.32
    áĢº
    -1.31
     flav
    -1.30
    POSITIVE LOGITS
    ĨĴ
    3.11
    Ī
    3.06
    Ļ
    3.02
    ķ
    2.91
    ľĵ
    2.85
    ĭ
    2.85
    Īĺ
    2.77
    ĻĤ
    2.75
    Ĵ
    2.75
    Ļª
    2.74
    Activations Density 0.182%

    No Known Activations