© 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. 13264
    Prev
    Next
    INDEX
    Explanations

    punctuation marks or periods indicating the end of 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/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
    +0.17
    1.0%
    258
    +0.14
    0.8%
    374
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    258
    +0.17
    0.03
    374
    +0.14
    0.02
    419
    +0.12
    0.02
    Negative Logits
    verte
    -1.53
    /,
    -1.43
    \.
    -1.43
     wont
    -1.41
     plating
    -1.37
    /)
    -1.37
    reb
    -1.36
    rente
    -1.35
     weekly
    -1.33
    plant
    -1.28
    POSITIVE LOGITS
    ĥ½
    2.00
    ¹
    1.94
    Ī
    1.88
    ģ
    1.86
    ½
    1.83
    Ģ
    1.81
    ĨĴ
    1.80
    Ĭ
    1.78
    ĥ
    1.76
    ī
    1.74
    Activations Density 0.095%

    No Known Activations