© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 10-RES-JB
    11. 7950
    Prev
    Next
    INDEX
    Explanations

    adjectives related to performance quality

    oai_token-act-pair · gpt-3.5-turbo

    descriptors associated with performance and achievement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Ingredients
    -0.79
     urgently
    -0.79
    ulhu
    -0.78
    $$$$
    -0.76
    conom
    -0.73
    actionDate
    -0.72
    galitarian
    -0.72
    helle
    -0.71
    skirts
    -0.71
    uphem
    -0.71
    POSITIVE LOGITS
     outing
    1.29
     performance
    1.12
     performances
    1.10
     finish
    1.09
     streak
    1.08
     matchup
    1.01
     outings
    1.00
     season
    0.98
     turnout
    0.97
     stint
    0.97
    Activations Density 0.148%

    No Known Activations