INDEX

Explanations

words related to downward motion or decrease

oai_token-act-pair · gpt-3.5-turbo

phrases related to declining trends or decreases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

cius

-0.77

thodox

-0.76

se

-0.73

ledge

-0.71

ific

-0.68

ioch

-0.66

iliary

-0.66

cyclopedia

-0.66

sic

-0.65

ilation

-0.65

POSITIVE LOGITS

BuyableInstoreAndOnline

0.95

 Falling

0.84

 prey

0.83

 asleep

0.79

fall

0.77

owship

0.77

ãĥĥãĥī

0.74

 graphene

0.74

 falling

0.73

 losses

0.69

Activations Density 0.013%

words related to downward motion or decrease

phrases related to declining trends or decreases

No Comments

No Known Activations