© 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. 4-RES-SM
    11. 9055
    Prev
    Next
    INDEX
    Explanations

    instances of conjunctions and phrases indicating transitions or contrasts in thought

    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
    255
    +0.14
    0.8%
    271
    +0.12
    0.7%
    316
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.14
    0.07
    181
    +0.12
    0.08
    143
    +0.11
    0.01
    Negative Logits
    atically
    -2.40
    izations
    -2.12
    matically
    -2.11
    ively
    -2.09
    istically
    -1.97
     instance
    -1.92
    isations
    -1.72
     example
    -1.68
    ically
    -1.61
    ized
    -1.60
    POSITIVE LOGITS
    Ļ
    2.79
    ĥ½
    2.67
    Ħ
    2.63
    ®
    2.60
    ħ
    2.54
    Ń
    2.32
    ī
    2.26
    ´
    2.26
    ±
    2.24
    ½
    2.22
    Activations Density 0.596%

    No Known Activations