© 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. 14148
    Prev
    Next
    INDEX
    Explanations

    negatively charged emotional words and phrases

    oai_token-act-pair · gpt-3.5-turbo

    emotional expressions and reactions to experiences

    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
    anwhile
    -0.79
    vati
    -0.74
    autions
    -0.72
    istrates
    -0.69
     Alternatively
    -0.66
    ":""},{"
    -0.65
    agher
    -0.64
    Alternatively
    -0.62
     disbanded
    -0.61
     Virtue
    -0.61
    POSITIVE LOGITS
     unbelievable
    1.60
     incredible
    1.47
     amazing
    1.39
     astounding
    1.35
     phenomenal
    1.32
     breathtaking
    1.28
     unbeliev
    1.24
     unreal
    1.22
    htaking
    1.21
     extraordinary
    1.18
    Activations Density 0.685%

    No Known Activations