© 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
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 12361
    Prev
    Next
    INDEX
    Explanations

    references to the material cotton and its various forms in different contexts

    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
    156
    +0.24
    1.4%
    230
    +0.12
    0.7%
    147
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    458
    +0.24
    0.30
    156
    +0.12
    0.29
    71
    +0.12
    0.28
    Negative Logits
    ths
    -1.84
    room
    -1.62
    bital
    -1.62
    ylase
    -1.46
    h
    -1.45
     precious
    -1.42
    gor
    -1.41
    iler
    -1.40
     far
    -1.40
    har
    -1.37
    POSITIVE LOGITS
    £
    3.20
    Ĵ
    2.75
    ł
    2.64
    ª
    2.52
    ¢
    2.27
    ĵ
    2.24
    ´
    2.22
    ¸
    2.19
    ¤
    2.17
    ĨĴ
    2.14
    Activations Density 1.412%

    No Known Activations