© 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. 30102
    Prev
    Next
    INDEX
    Explanations

    phrases that explain causes or motivations

    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
    5
    +0.13
    0.7%
    337
    +0.12
    0.7%
    397
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    61
    +0.13
    0.03
    5
    +0.12
    0.03
    372
    +0.12
    0.02
    Negative Logits
    ¿½
    -3.59
    ħ
    -3.34
    Ŀ
    -3.28
    ľ
    -3.21
    ¾
    -3.17
    ı
    -3.17
    Īĺ
    -3.14
    ¶
    -3.10
    ĨĴ
    -3.08
    ĵ
    -3.03
    POSITIVE LOGITS
     why
    2.41
    ableness
    2.30
    why
    1.77
    aways
    1.64
    itious
    1.59
    standing
    1.57
    ality
    1.55
    away
    1.54
     wait
    1.54
     driving
    1.52
    Activations Density 0.128%

    No Known Activations