© 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. 4-RES-SM
    11. 14253
    Prev
    Next
    INDEX
    Explanations

    contexts related to difficulty or challenges

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    28
    +0.14
    0.8%
    231
    +0.11
    0.6%
    78
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    28
    +0.14
    0.02
    485
    +0.11
    0.02
    502
    +0.11
    0.03
    Negative Logits
     entirety
    -1.93
    ¼
    -1.83
    Affirmed
    -1.74
    ¸
    -1.59
     expense
    -1.53
     proceeds
    -1.49
     classics
    -1.43
     preponderance
    -1.43
    ி
    -1.41
    gone
    -1.40
    POSITIVE LOGITS
     figuring
    2.56
     finding
    2.23
     locating
    2.18
     installing
    2.17
     accessing
    2.16
     seeing
    2.15
     parsing
    2.13
     converting
    2.08
     adjusting
    2.08
     getting
    2.08
    Activations Density 0.213%

    No Known Activations