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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 1-RES-SM
    6. 27104
    Prev
    Next
    INDEX
    Explanations

    possessive forms, particularly the word "’s" and its variants

    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
    478
    +0.15
    0.9%
    505
    +0.14
    0.8%
    14
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    405
    +0.15
    0.09
    505
    +0.14
    0.07
    128
    +0.11
    0.07
    Negative Logits
    ļ
    -1.76
    ups
    -1.66
     whatsoever
    -1.60
    ¢
    -1.60
    osa
    -1.58
    µ
    -1.50
    ģ
    -1.50
     unto
    -1.48
    ↵			
    -1.48
    ľ
    -1.48
    POSITIVE LOGITS
     been
    1.70
     birthday
    1.65
    rien
    1.65
    Decided
    1.64
     strongest
    1.53
     motto
    1.52
     Syndrome
    1.52
     reputation
    1.50
    been
    1.48
     greatest
    1.47
    Activations Density 0.448%

    No Known Activations