© 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. 4-RES-SM
    11. 11557
    Prev
    Next
    INDEX
    Explanations

    expressions of affirmations or confirmations

    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
    348
    +0.25
    1.5%
    21
    +0.15
    0.9%
    306
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    468
    +0.25
    0.04
    269
    +0.15
    0.04
    343
    +0.12
    0.03
    Negative Logits
    noreply
    -1.40
     bit
    -1.35
     lot
    -1.33
    sten
    -1.31
    ologically
    -1.31
    ately
    -1.30
    ually
    -1.29
     bunch
    -1.26
    bing
    -1.26
    hesis
    -1.25
    POSITIVE LOGITS
     as
    2.89
    as
    1.69
     como
    1.35
    below
    1.33
    еÐ
    1.31
    ovir
    1.26
    endor
    1.25
    ovirus
    1.24
    anine
    1.23
    ongs
    1.20
    Activations Density 0.095%

    No Known Activations