© 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. 28479
    Prev
    Next
    INDEX
    Explanations

    terms associated with locations and geographical references

    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
    352
    +0.15
    0.8%
    86
    +0.12
    0.7%
    191
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.15
    0.11
    18
    +0.12
    0.11
    86
    +0.12
    0.11
    Negative Logits
    aliana
    -1.72
    rapeut
    -1.69
    "));
    -1.48
    msgstr
    -1.48
     ."
    -1.48
     afore
    -1.45
    >",
    -1.44
    aintiff
    -1.43
     BASIS
    -1.41
    AFFIRMED
    -1.41
    POSITIVE LOGITS
    ´
    2.49
    ī
    2.44
    Ļ
    2.37
    ®
    2.30
    ·
    2.27
    ŀ
    2.24
    ¡
    2.24
    ¤
    2.24
    ¿
    2.23
    ¹
    2.22
    Activations Density 1.128%

    No Known Activations