INDEX

Explanations

phrases related to expressing strong opinions or emotions

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

 seiz

-0.81

 sacrific

-0.71

illary

-0.71

eatures

-0.69

leneck

-0.69

rys

-0.68

luaj

-0.67

glomer

-0.67

ixel

-0.66

successfully

-0.65

POSITIVE LOGITS

ifiable

1.24

 plain

1.08

ifications

0.93

 coincidence

0.81

 kidding

0.80

 icing

0.79

 deserts

0.74

 desserts

0.73

 gonna

0.73

IFIED

0.72

Activations Density 0.047%

phrases related to expressing strong opinions or emotions

No Comments

No Known Activations