© 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. 3-RES-JB
    11. 11289
    Prev
    Next
    INDEX
    Explanations

    default settings or values in different contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to default settings or configurations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    borough
    -0.94
    ropolitan
    -0.82
    atform
    -0.82
    kefeller
    -0.81
    chn
    -0.76
    hemat
    -0.74
    areth
    -0.74
    aph
    -0.74
    ales
    -0.73
    andan
    -0.72
    POSITIVE LOGITS
     defaults
    1.11
     settings
    0.98
     default
    0.91
     behaviour
    0.86
     behavior
    0.81
     gateway
    0.81
     configuration
    0.79
    dict
    0.75
     exposure
    0.75
     implementations
    0.74
    Activations Density 0.009%

    No Known Activations