© 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. 27924
    Prev
    Next
    INDEX
    Explanations

    repetitive phrases that express greetings or welcomes

    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
    156
    +0.25
    1.4%
    410
    +0.17
    1.0%
    376
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.25
    0.13
    156
    +0.17
    0.07
    271
    +0.15
    0.06
    Negative Logits
    ®
    -2.44
    §
    -2.40
    ĥ½
    -2.27
    ¯
    -2.23
    Ń
    -2.15
    ³
    -2.12
    ij
    -2.11
    ļ
    -2.09
    ´
    -2.07
    ¹
    -2.05
    POSITIVE LOGITS
     familiar
    1.49
     seeking
    1.48
    ward
    1.43
     volunteers
    1.41
    atus
    1.37
    gers
    1.34
     ask
    1.33
     sound
    1.31
     seekers
    1.30
     Posts
    1.29
    Activations Density 0.392%

    No Known Activations