INDEX

Explanations

phrases indicating a level of involvement or connection to a situation or idea

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate actions or experiences in a casual or conversational context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 urgently

-0.70

 proudly

-0.68

 purportedly

-0.66

 reportedly

-0.65

 ostensibly

-0.64

 routinely

-0.64

 celebrated

-0.63

 widely

-0.62

 preferably

-0.61

ewitness

-0.61

POSITIVE LOGITS

ahime

0.93

onite

0.80

tin

0.78

kered

0.67

paralle

0.67

icable

0.66

appro

0.65

tle

0.65

 misunder

0.64

 glim

0.63

Activations Density 0.433%

phrases indicating a level of involvement or connection to a situation or idea

phrases that indicate actions or experiences in a casual or conversational context

No Comments

No Known Activations