© 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. 612
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "achers" which seems to have been cropped from the word "poachers."

    oai_token-act-pair · gpt-3.5-turbo

    mentions of teachers

    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
     cycle
    -0.66
     sanity
    -0.65
     substance
    -0.65
     flaw
    -0.64
     Substance
    -0.63
     Cycle
    -0.63
     shorth
    -0.62
     continuity
    -0.62
     PID
    -0.61
     procedure
    -0.61
    POSITIVE LOGITS
    achers
    4.30
    acher
    2.03
    aching
    1.97
    aches
    1.80
    ached
    1.42
    achment
    1.29
    akers
    1.24
    earchers
    1.22
    ammers
    1.20
    ACH
    1.15
    Activations Density 0.014%

    No Known Activations