© 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. 30233
    Prev
    Next
    INDEX
    Explanations

    data structures and database-related terms

    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
    +0.18
    1.0%
    123
    +0.12
    0.7%
    330
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    157
    +0.18
    0.16
    330
    +0.12
    0.08
    224
    +0.11
    0.03
    Negative Logits
    sciously
    -1.62
    agus
    -1.49
    awning
    -1.46
    inally
    -1.41
    shit
    -1.40
    ajÄħ
    -1.38
    OOST
    -1.36
    ermine
    -1.34
    icate
    -1.34
    essen
    -1.34
    POSITIVE LOGITS
    ĭ
    6.48
    ĻĤ
    6.42
    Ļ
    6.40
    ħ
    6.30
    ĥ½
    6.23
    Ī
    6.22
    ¹
    6.21
    ĩ
    6.20
    ¿½
    6.16
    ª
    6.16
    Activations Density 4.576%

    No Known Activations