© 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. 12581
    Prev
    Next
    INDEX
    Explanations

    instances of the word "look" and its variations, indicating a focus on visual perception or attention

    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
    339
    +0.13
    0.7%
    164
    +0.13
    0.7%
    396
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    260
    +0.13
    0.04
    45
    +0.13
    0.04
    339
    +0.13
    0.04
    Negative Logits
    ŀ
    -5.25
    Ī
    -5.22
    §
    -5.14
    ļ
    -4.87
    į
    -4.87
    Ł
    -4.74
    Į
    -4.71
    ĭ
    -4.59
    ī
    -4.52
    ĥ
    -4.50
    POSITIVE LOGITS
    ahead
    1.68
     inward
    1.60
    int
    1.59
    lon
    1.54
    dad
    1.50
     0000000000000000000000000000000000
    1.44
    henyl
    1.43
    RSOS
    1.35
    backs
    1.33
    gay
    1.33
    Activations Density 0.123%

    No Known Activations