© 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. 4-RES-SM
    11. 10424
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing the quality or nature of experiences and objects

    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
    24
    +0.15
    0.8%
    328
    +0.12
    0.7%
    144
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    24
    +0.15
    0.04
    137
    +0.12
    0.06
    436
    +0.11
    0.03
    Negative Logits
    ĨĴ
    -3.10
    ¥
    -3.04
    Ĥ
    -3.01
    ¢
    -3.01
    »
    -3.00
    »¿
    -3.00
    ł
    -2.96
    ĥ½
    -2.96
    ŀ
    -2.88
    ī
    -2.87
    POSITIVE LOGITS
     basis
    2.41
     footing
    2.24
     behalf
    2.19
     whim
    2.11
     pedest
    2.08
     occasions
    1.95
     canvas
    1.94
     scale
    1.85
    usp
    1.79
     pretext
    1.74
    Activations Density 0.502%

    No Known Activations