© 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. 20056
    Prev
    Next
    INDEX
    Explanations

    references to specific names, potentially related to pride or familial ties

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of specific individuals and terms related to productivity and output

    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
     Psy
    -0.86
     Emin
    -0.86
    Ĭ±
    -0.85
    Psy
    -0.81
    Aust
    -0.80
     Ãľ
    -0.77
     Vienna
    -0.77
    ür
    -0.76
    Steam
    -0.75
     psycho
    -0.73
    POSITIVE LOGITS
     Cain
    2.50
     Cull
    1.87
     Clayton
    1.46
     Holl
    1.45
     cull
    1.28
     Jindal
    1.26
     Boone
    1.17
     churn
    1.16
     Boll
    1.12
     Cutter
    1.11
    Activations Density 0.039%

    No Known Activations