© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 1-RES-SM
    11. 26287
    Prev
    Next
    INDEX
    Explanations

    the word "which" indicating relative clauses or additional information in sentences

    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
    +0.14
    0.8%
    70
    +0.12
    0.7%
    144
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    319
    +0.14
    0.04
    262
    +0.12
    0.04
    248
    +0.11
    0.04
    Negative Logits
    lord
    -1.45
    mine
    -1.37
    remember
    -1.37
    être
    -1.36
    cav
    -1.35
    raising
    -1.35
     Ded
    -1.32
    sequence
    -1.30
     Wrote
    -1.30
    :**
    -1.29
    POSITIVE LOGITS
    į
    2.81
    Į
    2.60
    ĥ½
    2.50
    ı
    2.44
    ¹
    2.42
    ħ
    2.39
    IJ
    2.37
    İ
    2.35
    Ĵ
    2.34
    ŀ
    2.28
    Activations Density 0.131%

    No Known Activations