© 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. 13290
    Prev
    Next
    INDEX
    Explanations

    legal terminology and references to participants in legal cases

    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
    410
    +0.18
    1.0%
    351
    +0.14
    0.8%
    80
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.18
    0.03
    80
    +0.14
    0.02
    365
    +0.12
    0.02
    Negative Logits
     ago
    -1.72
     technologies
    -1.58
     enriched
    -1.56
     immersion
    -1.45
     enrichment
    -1.44
    />
    -1.40
     fois
    -1.40
    ashing
    -1.37
    lication
    -1.32
    iring
    -1.31
    POSITIVE LOGITS
    ĻĤ
    2.92
    ¥
    2.67
    ±
    2.66
    ĸ´
    2.48
    Īĺ
    2.45
    ¿
    2.31
    ³
    2.27
    ²
    2.22
    £
    2.21
    ĵ
    2.21
    Activations Density 0.080%

    No Known Activations