© 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. 1-RES-JB
    11. 4327
    Prev
    Next
    INDEX
    Explanations

    strings of characters that seem to be from a separate layer in a neural network or artificial intelligence system

    oai_token-act-pair · gpt-3.5-turbo

    emotional and descriptive language related to performance or artistic expression

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    hement
    -1.18
     veter
    -0.95
    ĸļ
    -0.93
     ende
    -0.90
     tremend
    -0.89
     destro
    -0.89
    ulkan
    -0.87
    ername
    -0.87
     mathemat
    -0.81
     eleph
    -0.81
    POSITIVE LOGITS
    âĹı
    0.82
    Spain
    0.76
    Var
    0.74
    Mand
    0.74
    SIM
    0.73
    âĺħ
    0.72
    ³³³³³³³³³³³³³³³³
    0.71
    Reply
    0.71
    Ba
    0.71
    Moon
    0.71
    Activations Density 0.745%

    No Known Activations