Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 3360
    Prev
    Next
    INDEX
    Explanations

    sentences where the word "But" is used to contrast different parts of the text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "But" used to introduce contrasting information

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    SourceFile
    -0.88
    MpServer
    -0.70
    interstitial
    -0.70
    NRS
    -0.65
    Pool
    -0.65
    olves
    -0.65
     Coverage
    -0.62
    ãĤ´
    -0.60
    ļéĨĴ
    -0.60
    clave
    -0.59
    POSITIVE LOGITS
     alas
    1.20
     yeah
    0.96
     uh
    0.95
     interestingly
    0.93
     secondly
    0.89
     yes
    0.89
     unlike
    0.88
     hey
    0.88
     cru
    0.85
     moreover
    0.85
    Activations Density 0.090%

    No Known Activations