INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals' names

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

flies

-0.87

enture

-0.78

ãĥĺãĥ©

-0.73

ioned

-0.73

doors

-0.72

OTE

-0.72

fly

-0.72

nel

-0.71

ledged

-0.71

bone

-0.70

POSITIVE LOGITS

ahon

0.79

 thickness

0.76

zzi

0.74

ucci

0.68

otti

0.67

borough

0.66

oples

0.66

Dru

0.65

acht

0.64

ueller

0.63

Activations Density 0.033%

names of individuals

mentions of specific individuals' names

No Comments

No Known Activations