© 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. 0-RES-JB
    11. 20407
    Prev
    Next
    INDEX
    Explanations

    the word "throng" in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "thrash" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    isSpecialOrderable
    -0.78
     Kee
    -0.69
    etting
    -0.69
    ogenesis
    -0.67
     iOS
    -0.66
    ridge
    -0.65
    dar
    -0.65
    embed
    -0.65
     macOS
    -0.64
    Creating
    -0.63
    POSITIVE LOGITS
     thr
    4.06
     Thr
    1.97
     clut
    1.05
     THR
    0.97
     thrill
    0.97
     thro
    0.97
     trou
    0.97
     Thrust
    0.91
     fle
    0.88
     whim
    0.87
    Activations Density 0.006%

    No Known Activations