© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29684
    Prev
    Next
    INDEX
    Explanations

    mentions of profiles and profile-related attributes

    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
    386
    +0.15
    0.9%
    156
    +0.14
    0.8%
    255
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    269
    +0.15
    0.02
    38
    +0.14
    0.01
    46
    +0.14
    0.01
    Negative Logits
    ´
    -4.57
    º
    -4.40
    ĥ½
    -4.09
    ŀ
    -3.99
    §
    -3.97
    Ĵ
    -3.97
    Ł
    -3.95
    ¤
    -3.89
    ª
    -3.86
    °
    -3.77
    POSITIVE LOGITS
    pective
    1.73
    omen
    1.47
    rett
    1.46
    ionate
    1.44
     Info
    1.44
    ew
    1.43
    igraph
    1.43
    holder
    1.43
    ylene
    1.42
    ically
    1.42
    Activations Density 0.155%

    No Known Activations