© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 11041
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing intensity or significance in a variety of contexts

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    125
    +0.12
    0.7%
    8
    +0.12
    0.7%
    45
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    289
    +0.12
    0.01
    133
    +0.12
    0.03
    45
    +0.12
    0.02
    Negative Logits
    ±
    -4.87
    µ
    -4.86
    »¿
    -4.83
    °
    -4.77
    Į
    -4.67
    ij
    -4.60
    ¥
    -4.59
    £
    -4.54
    ĭ
    -4.53
    ¯
    -4.52
    POSITIVE LOGITS
     over
    3.62
     Over
    3.00
    over
    2.94
    Over
    2.70
     OVER
    2.58
    overs
    2.42
     under
    2.34
     overl
    2.27
    under
    2.27
     Under
    2.21
    Activations Density 0.471%

    No Known Activations