INDEX

Explanations

phrases related to neglectful behavior and potentially harmful situations

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

CHAT

-0.74

 VIDE

-0.74

Streamer

-0.70

SW

-0.63

Rated

-0.63

soType

-0.62

isse

-0.61

Flo

-0.60

night

-0.60

 Aires

-0.60

POSITIVE LOGITS

ful

0.97

fully

0.88

fulness

0.83

FUL

0.78

shire

0.75

icit

0.74

reatment

0.73

WARE

0.73

luster

0.72

mental

0.71

Activations Density 0.046%

phrases related to neglectful behavior and potentially harmful situations

No Comments

No Known Activations