© 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. 10955
    Prev
    Next
    INDEX
    Explanations

    speculative language and expressions of uncertainty

    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
    +0.30
    1.8%
    25
    +0.20
    1.2%
    174
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.30
    0.09
    361
    +0.20
    0.07
    330
    +0.10
    0.08
    Negative Logits
    į
    -3.91
    ĥ
    -3.68
    Ģ
    -3.56
    ħ
    -3.55
    ľĵ
    -3.45
    «
    -3.44
    ¢
    -3.44
    Ĥ
    -3.43
    Ħ
    -3.43
    ¿
    -3.29
    POSITIVE LOGITS
     it
    2.31
     there
    2.30
     we
    2.17
     somebody
    2.06
     they
    2.06
     maybe
    2.03
     someone
    2.02
     you
    2.01
     someday
    1.96
     thats
    1.92
    Activations Density 0.953%

    No Known Activations