INDEX

Explanations

expressions related to emotional reactions or states

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 festive

-0.69

 sugg

-0.69

ãĥĩãĤ£

-0.67

ãĥ©ãĥ³

-0.63

xtap

-0.62

coh

-0.59

ãĥĹ

-0.57

å¼

-0.56

Points

-0.56

 Branch

-0.56

POSITIVE LOGITS

!--

0.83

 Himself

0.78

 anyways

0.74

nesses

0.73

 anyway

0.72

selves

0.72

 ALSO

0.72

 nonetheless

0.71

 insofar

0.71

 WHEN

0.69

Activations Density 0.478%

expressions related to emotional reactions or states

No Comments

No Known Activations