© 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. 4-RES-JB
    11. 13049
    Prev
    Next
    INDEX
    Explanations

    expressions of surprise or confusion

    oai_token-act-pair · gpt-3.5-turbo

    punctuation that expresses surprise or confusion

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

    No Comments

    Negative Logits
    itions
    -0.81
    Ö¼
    -0.77
    atives
    -0.77
    ©¶æ
    -0.71
    iatric
    -0.70
    opath
    -0.69
    oran
    -0.68
    atri
    -0.67
    ridor
    -0.67
    oreal
    -0.67
    POSITIVE LOGITS
    ?!
    1.25
    #$
    0.92
     Huh
    0.89
    ?,
    0.88
    @#
    0.88
    @#&
    0.83
     Seriously
    0.82
    ?:
    0.82
    !!
    0.81
     Why
    0.80
    Activations Density 0.012%

    No Known Activations