INDEX

Explanations

phrases related to direct confrontations or interactions

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a confrontation or interaction

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Penal

-0.63

hower

-0.60

 contemplated

-0.59

ONEY

-0.57

Mb

-0.57

©¶æ¥µ

-0.57

Ri

-0.57

KEY

-0.56

 banners

-0.56

 Fowler

-0.56

POSITIVE LOGITS

vised

0.89

lihood

0.87

alog

0.83

quartered

0.77

etheless

0.75

wards

0.75

alogue

0.73

ãĤ¨ãĥ«

0.70

mination

0.69

ership

0.67

Activations Density 0.074%

phrases related to direct confrontations or interactions

phrases that indicate a confrontation or interaction

No Comments

No Known Activations