© 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. 14638
    Prev
    Next
    INDEX
    Explanations

    references to figures or tables in the text

    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
    255
    +0.14
    0.8%
    78
    +0.13
    0.7%
    306
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.14
    0.00
    78
    +0.13
    0.00
    388
    +0.12
    0.01
    Negative Logits
    ities
    -1.74
     chance
    -1.55
    ologist
    -1.48
    ubicin
    -1.46
    manship
    -1.41
     incumbent
    -1.37
    atically
    -1.37
    caster
    -1.37
    arity
    -1.36
    iginally
    -1.34
    POSITIVE LOGITS
    ²
    3.74
    ŀ
    3.70
    ¸
    3.69
    ĩ
    3.67
    ¢
    3.65
    Ī
    3.61
    į
    3.54
    ĥ
    3.54
    ¶
    3.51
    Ĥ¬
    3.48
    Activations Density 0.053%

    No Known Activations