© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 15414
    Prev
    Next
    INDEX
    Explanations

    phrases indicating what someone thinks or believes

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "thinks" and its variations

    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
    bearing
    -0.66
     completion
    -0.61
     approximately
    -0.61
     total
    -0.61
     precip
    -0.60
     mounting
    -0.59
     spring
    -0.59
     televised
    -0.59
     stage
    -0.58
     phase
    -0.58
    POSITIVE LOGITS
     thinks
    3.05
     believes
    2.14
     considers
    1.95
     expects
    1.85
     knows
    1.72
     understands
    1.70
     realizes
    1.69
     disagrees
    1.69
     remembers
    1.68
     prefers
    1.66
    Activations Density 0.031%

    No Known Activations