INDEX

Explanations

phrases related to political figures, policies, and international affairs

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

flush

-0.63

taboola

-0.62

ãĥĦ

-0.61

ewater

-0.61

wash

-0.59

wa

-0.58

cov

-0.57

eland

-0.54

æ©Ł

-0.54

\'

-0.54

POSITIVE LOGITS

 what

0.92

how

0.79

those

0.72

 whether

0.71

what

0.70

whether

0.68

 those

0.68

 some

0.67

acts

0.64

ample

0.63

Activations Density 1.518%

phrases related to political figures, policies, and international affairs

No Comments

No Known Activations