© 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. 1-RES-JB
    11. 4880
    Prev
    Next
    INDEX
    Explanations

    phrases related to deception or manipulation, often involving the term "bull"

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "bull" or its variations, indicating content related to market dynamics or aggressive discourse

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Interstitial
    -0.74
    ONS
    -0.73
    ALLY
    -0.71
    اÙĦ
    -0.70
    Ô
    -0.70
    ATIVE
    -0.67
    ATING
    -0.65
    USS
    -0.65
    OURCE
    -0.64
     Designer
    -0.64
    POSITIVE LOGITS
    bull
    1.33
    scl
    0.98
    shit
    0.98
    enstein
    0.94
    aroo
    0.94
    wagen
    0.92
    emouth
    0.88
    dogs
    0.86
    emort
    0.86
    ertodd
    0.84
    Activations Density 0.002%

    No Known Activations