© 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. 711
    Prev
    Next
    INDEX
    Explanations

    words ending with the suffix "-icious"

    oai_token-act-pair · gpt-4-turbo

    words related to maliciousness or suspiciousness

    oai_token-act-pair · gpt-3.5-turbo

    adjective forms of words that imply negativity or suspicion

    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
    mberg
    -0.85
    ETA
    -0.76
    uph
    -0.68
    udeb
    -0.67
    ©¶æ
    -0.66
    locked
    -0.66
    orph
    -0.65
    puter
    -0.65
    oling
    -0.65
     Skydragon
    -0.64
    POSITIVE LOGITS
    icious
    1.33
    ly
    1.07
    nesses
    1.01
    Magikarp
    0.96
    NESS
    0.88
    ness
    0.86
    ãĥ¼ãĥĨãĤ£
    0.86
    ively
    0.84
    ortun
    0.81
     Intent
    0.81
    Activations Density 0.020%

    No Known Activations