© 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. 14693
    Prev
    Next
    INDEX
    Explanations

    specific names and terms related to people and organizations

    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
    307
    +0.17
    0.9%
    9
    +0.16
    0.9%
    146
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.17
    0.07
    181
    +0.16
    0.08
    285
    +0.12
    0.08
    Negative Logits
    izable
    -2.07
    izations
    -1.86
     heartbeat
    -1.81
    imately
    -1.70
    ://
    -1.70
     presence
    -1.56
    irtual
    -1.44
     other
    -1.44
     FIG
    -1.44
     utilization
    -1.43
    POSITIVE LOGITS
    ·
    4.63
    º
    4.53
    ¿
    4.51
    »¿
    4.49
    ®
    4.44
    Ļª
    4.41
    Ļ
    4.38
    µ
    4.35
    ª
    4.30
    İ
    4.29
    Activations Density 0.423%

    No Known Activations