© 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. 31923
    Prev
    Next
    INDEX
    Explanations

    various forms of the word "for."

    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
    +0.21
    1.2%
    97
    +0.13
    0.8%
    156
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.21
    0.13
    310
    +0.13
    0.09
    478
    +0.12
    0.02
    Negative Logits
    stra
    -1.75
     because
    -1.71
     racket
    -1.60
    \[[@
    -1.54
    )](#
    -1.50
    PA
    -1.50
     but
    -1.48
    because
    -1.46
    ful
    -1.45
    lich
    -1.45
    POSITIVE LOGITS
    ľĵ
    3.03
    ¿½
    2.82
    ı
    2.63
    ¬
    2.60
    ĻĤ
    2.59
    Īĺ
    2.55
    ĺ
    2.54
    ®
    2.47
    IJ
    2.45
    Ĭ
    2.44
    Activations Density 2.130%

    No Known Activations