© 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. 1-RES-SM
    11. 23854
    Prev
    Next
    INDEX
    Explanations

    instances of the word "apparently" and other similar qualifiers indicating uncertainty or subjective assessments

    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/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
    Neuron Alignment
    Index
    Value
    % of L₁
    23
    +0.28
    1.6%
    376
    +0.16
    0.9%
    377
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    377
    +0.28
    0.04
    78
    +0.16
    0.03
    74
    +0.12
    0.03
    Negative Logits
    Ģ
    -3.27
    ©
    -3.21
    ĭ
    -3.19
    §
    -3.11
    Ķ
    -3.09
    Ĭ
    -3.08
    Ĵ
    -3.00
    Ĥ¬
    -2.98
    Ħ
    -2.97
    £
    -2.95
    POSITIVE LOGITS
     unemploy
    1.82
     tranqu
    1.72
     suicidal
    1.66
     '')
    1.55
     ammonium
    1.55
     inco
    1.50
     halluc
    1.49
     harmless
    1.48
     unable
    1.45
     vomiting
    1.44
    Activations Density 0.120%

    No Known Activations