© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 30297
    Prev
    Next
    INDEX
    Explanations

    phrases indicating personal experiences and reflections

    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
    358
    +0.12
    0.7%
    302
    +0.11
    0.6%
    244
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    88
    +0.12
    0.15
    358
    +0.11
    0.16
    108
    +0.10
    0.14
    Negative Logits
     unfolded
    -1.82
    tees
    -1.42
     whom
    -1.40
    zzi
    -1.39
     reflections
    -1.35
    ring
    -1.34
     himself
    -1.33
     light
    -1.32
     rounds
    -1.32
     Papers
    -1.31
    POSITIVE LOGITS
    they
    1.60
     extinct
    1.54
     lovers
    1.53
     lately
    1.51
     they
    1.48
    arcin
    1.46
    winning
    1.43
     dating
    1.40
    :&
    1.37
    rue
    1.34
    Activations Density 2.294%

    No Known Activations