© 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. 1-RES-SM
    11. 27780
    Prev
    Next
    INDEX
    Explanations

    specific formatting and syntax related to code structure, particularly in user interface design

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    250
    +0.15
    0.9%
    478
    +0.13
    0.7%
    380
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    250
    +0.15
    0.04
    380
    +0.13
    0.02
    162
    +0.11
    0.02
    Negative Logits
     ](
    -1.84
    inguished
    -1.71
    omic
    -1.66
    ...](
    -1.54
    omorphism
    -1.54
    aea
    -1.52
    omial
    -1.49
    omal
    -1.47
    rise
    -1.45
    itious
    -1.44
    POSITIVE LOGITS
    xico
    1.51
     linger
    1.50
     fuck
    1.40
     aside
    1.40
     antagonist
    1.39
    %-
    1.38
     bras
    1.35
     cele
    1.35
    tras
    1.34
    +)
    1.34
    Activations Density 0.182%

    No Known Activations