© 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. 2-RES-SM
    11. 32036
    Prev
    Next
    INDEX
    Explanations

    code-related structures and definitions

    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
    72
    +0.23
    1.3%
    198
    +0.16
    0.9%
    186
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    72
    +0.23
    0.13
    85
    +0.16
    0.05
    86
    +0.15
    0.10
    Negative Logits
    blast
    -1.72
    igation
    -1.62
    oses
    -1.49
    ansion
    -1.44
    bos
    -1.43
    projects
    -1.41
    ()</
    -1.37
    fills
    -1.35
    ismic
    -1.34
    Argued
    -1.34
    POSITIVE LOGITS
    ¹
    1.60
    ¸
    1.57
     whom
    1.55
     affiliate
    1.53
     Nielsen
    1.48
    ¬
    1.44
     Morgan
    1.42
    âģĦ
    1.41
     Virgin
    1.37
     dow
    1.36
    Activations Density 4.708%

    No Known Activations