© 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 EvalsExportsinterp-engineNEWAPI 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. 13359
    Prev
    Next
    INDEX
    Explanations

    references to judges and judicial opinions

    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
    407
    +0.14
    0.8%
    392
    +0.12
    0.7%
    21
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    89
    +0.14
    0.02
    177
    +0.12
    0.02
    168
    +0.12
    0.02
    Negative Logits
    aste
    -1.76
    msgstr
    -1.71
    ...](
    -1.66
     connection
    -1.63
    enic
    -1.51
    ed
    -1.50
    API
    -1.42
    \)
    -1.41
    jection
    -1.41
    connection
    -1.40
    POSITIVE LOGITS
     aloud
    2.00
     whom
    1.73
     himself
    1.73
     opinion
    1.72
     opinions
    1.68
     appointed
    1.59
    itti
    1.54
     Appeal
    1.44
    committee
    1.43
     excused
    1.42
    Activations Density 0.080%

    No Known Activations