© 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. 9624
    Prev
    Next
    INDEX
    Explanations

    phrases indicating exceptions and specific conditions

    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
    419
    +0.23
    1.4%
    186
    +0.22
    1.3%
    198
    +0.18
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    198
    +0.23
    0.06
    419
    +0.22
    0.06
    186
    +0.18
    0.04
    Negative Logits
    ¦
    -1.75
    ¡
    -1.71
    ¯
    -1.66
    Ļ
    -1.63
    ¨
    -1.59
    ¿½
    -1.58
    º
    -1.52
    Ļª
    -1.51
    ij
    -1.50
    ´
    -1.49
    POSITIVE LOGITS
    any
    1.57
     ground
    1.51
    nier
    1.48
     slightest
    1.44
     surprises
    1.44
    else
    1.41
    orable
    1.40
     except
    1.36
    awa
    1.34
    quist
    1.33
    Activations Density 0.781%

    No Known Activations