© 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. 2-RES-JB
    11. 19414
    Prev
    Next
    INDEX
    Explanations

    the word "Reason" with high activation values

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with reasoning and innovation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ardless
    -0.87
    arding
    -0.81
    iannopoulos
    -0.80
    ÑĢ
    -0.74
    uders
    -0.71
    olson
    -0.71
    è£
    -0.69
    Ïī
    -0.68
    å£
    -0.68
    uding
    -0.67
    POSITIVE LOGITS
     Matters
    1.22
     Solutions
    1.17
     Scores
    1.09
     Transfer
    1.09
     Mechanics
    1.06
     Solution
    1.05
     Points
    1.04
     Score
    1.04
     Problems
    1.00
     Pieces
    0.98
    Activations Density 0.089%

    No Known Activations