© 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. 10925
    Prev
    Next
    INDEX
    Explanations

    phrases indicating varying levels of disability or impairment

    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
    156
    +0.18
    1.0%
    59
    +0.12
    0.6%
    472
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    18
    +0.18
    0.01
    148
    +0.12
    0.00
    59
    +0.11
    0.01
    Negative Logits
    fulness
    -1.86
    ynes
    -1.54
    hav
    -1.50
    ène
    -1.45
    omb
    -1.44
    fully
    -1.42
    ually
    -1.42
    instein
    -1.41
     fond
    -1.41
    cles
    -1.38
    POSITIVE LOGITS
    ¿½
    1.82
    heses
    1.74
    Ĥ
    1.61
    ities
    1.57
    Ħ
    1.54
    ĻĤ
    1.53
    aneous
    1.52
     sized
    1.48
     least
    1.45
    µ
    1.43
    Activations Density 0.074%

    No Known Activations