© 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. 0-RES-JB
    11. 16958
    Prev
    Next
    INDEX
    Explanations

    instances of the word "speculate"

    oai_token-act-pair · gpt-3.5-turbo

    instances of speculation and discussions around ransom

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     conformity
    -0.70
     ven
    -0.70
    warts
    -0.67
    natureconservancy
    -0.67
     boundaries
    -0.65
    viol
    -0.65
    raped
    -0.65
    Viol
    -0.65
    videos
    -0.64
     swearing
    -0.64
    POSITIVE LOGITS
     speculate
    2.57
     speculated
    1.88
     ransom
    1.83
     Islanders
    1.42
     spec
    1.29
     idle
    1.27
     theor
    1.23
     speculative
    1.18
     hypothes
    1.11
     ferry
    1.09
    Activations Density 0.048%

    No Known Activations