© 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. 11102
    Prev
    Next
    INDEX
    Explanations

    specific attributes or characteristics related to data structures and programming

    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
    307
    +0.15
    0.9%
    132
    +0.12
    0.7%
    428
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    478
    +0.15
    0.02
    397
    +0.12
    0.04
    47
    +0.11
    0.05
    Negative Logits
    FA
    -1.38
    cule
    -1.36
    ogen
    -1.34
    ipers
    -1.31
     Fan
    -1.31
    ichi
    -1.28
    lig
    -1.28
    atal
    -1.26
     Tro
    -1.25
    imir
    -1.24
    POSITIVE LOGITS
     c
    3.28
     t
    2.91
     b
    2.77
     s
    2.68
     d
    2.68
     v
    2.64
     l
    2.57
     h
    2.56
     m
    2.56
     i
    2.54
    Activations Density 1.239%

    No Known Activations