© 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. 1-RES-SM
    11. 26507
    Prev
    Next
    INDEX
    Explanations

    references to the term "back" in various contexts

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    374
    +0.12
    0.7%
    477
    +0.10
    0.6%
    463
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    374
    +0.12
    0.05
    415
    +0.10
    0.04
    463
    +0.10
    0.04
    Negative Logits
    «
    -2.27
    ĵ
    -2.23
    §
    -2.22
    »¿
    -2.07
     fit
    -1.96
    ¯
    -1.90
    ĩ
    -1.84
    ¥
    -1.81
    Ļ
    -1.76
    ½
    -1.76
    POSITIVE LOGITS
    lund
    1.95
    woods
    1.87
    door
    1.82
    rÃŃ
    1.75
    tracking
    1.73
    crossing
    1.72
    breaking
    1.71
    enthal
    1.68
    lessly
    1.67
    stage
    1.58
    Activations Density 0.093%

    No Known Activations