© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 26082
    Prev
    Next
    INDEX
    Explanations

    terms related to property attributes and descriptors

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    100
    +0.15
    0.9%
    93
    +0.13
    0.7%
    467
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    93
    +0.15
    0.03
    467
    +0.13
    0.02
    100
    +0.12
    0.03
    Negative Logits
    ³
    -4.56
    Ń
    -4.55
    Ŀ
    -4.52
    ↵
    -4.45
    -4.45
    -4.45
    ↵
    -4.45
    <|outofrange|>
    -4.45
    -4.45
    -4.45
    POSITIVE LOGITS
    ious
    1.85
    Changed
    1.70
    itious
    1.61
    props
    1.59
     keys
    1.54
    eries
    1.51
    ality
    1.50
    etting
    1.50
     deteriorated
    1.48
    ivated
    1.48
    Activations Density 0.156%

    No Known Activations