© 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. 2-RES-JB
    11. 17760
    Prev
    Next
    INDEX
    Explanations

    words or phrases that are used frequently or repeatedly in a given context

    oai_token-act-pair · gpt-3.5-turbo

    repetitive phrases emphasizing frequency of occurrence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    utenberg
    -0.84
    kay
    -0.77
    agate
    -0.76
    gur
    -0.72
    gae
    -0.70
    shi
    -0.69
    plates
    -0.68
    ged
    -0.66
    zbek
    -0.65
    imming
    -0.65
    POSITIVE LOGITS
    entimes
    1.38
     overlooked
    0.97
     encountered
    0.91
     resorted
    0.91
     referred
    0.84
     misunderstood
    0.83
     cited
    0.83
    theless
    0.82
     relied
    0.81
     touted
    0.81
    Activations Density 0.034%

    No Known Activations