© 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. 27634
    Prev
    Next
    INDEX
    Explanations

    references to accounting and finance-related terms

    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
    63
    +0.17
    1.0%
    442
    +0.16
    0.9%
    161
    +0.15
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    63
    +0.17
    0.10
    442
    +0.16
    0.16
    49
    +0.15
    0.17
    Negative Logits
    tera
    -1.74
    /~
    -1.53
    enberg
    -1.51
    ))$.
    -1.40
    lei
    -1.36
    ))$
    -1.35
     quo
    -1.32
    rote
    -1.29
     Â
    -1.28
     thereof
    -1.25
    POSITIVE LOGITS
    ¤
    2.31
    £
    2.17
    Ī
    2.16
    Ĥ¬
    2.10
    IJ
    2.05
    ģ
    2.05
    Ĩ
    2.03
    ľĵ
    2.03
    ©
    1.99
    ŀ
    1.98
    Activations Density 7.717%

    No Known Activations