© 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. 11-RES-JB
    11. 18672
    Prev
    Next
    INDEX
    Explanations

    instances where something is not happening as expected or intended, and an alternative action is taken instead

    oai_token-act-pair · gpt-3.5-turbo

    instances of the term "instead" and its associated context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    not
    -0.76
    ilty
    -0.68
    NOT
    -0.67
    no
    -0.62
     NOT
    -0.62
     1886
    -0.61
    KO
    -0.61
     yes
    -0.61
     BEFORE
    -0.59
    enough
    -0.59
    POSITIVE LOGITS
    Instead
    1.16
     Instead
    1.09
     opted
    0.97
     instead
    0.96
    ocused
    0.90
     opting
    0.90
     concentrate
    0.89
     focused
    0.87
     focus
    0.87
    instead
    0.86
    Activations Density 0.245%

    No Known Activations