Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 12153
    Prev
    Next
    INDEX
    Explanations

    emotional reflections and personal growth themes

    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
    271
    +0.13
    0.7%
    225
    +0.12
    0.7%
    365
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.13
    0.11
    266
    +0.12
    0.10
    53
    +0.10
    0.11
    Negative Logits
    lass
    -1.89
    fered
    -1.54
     whom
    -1.53
    ities
    -1.41
     liability
    -1.41
    eded
    -1.37
    vendor
    -1.36
     interchange
    -1.33
     defendants
    -1.32
     frozen
    -1.31
    POSITIVE LOGITS
    Writ
    1.54
    blr
    1.51
     story
    1.51
    rae
    1.49
    ···
    1.45
    roc
    1.44
     apolog
    1.44
    ******
    1.40
    erta
    1.40
     worrying
    1.39
    Activations Density 1.265%

    No Known Activations