INDEX

Explanations

phrases related to recent years

oai_token-act-pair · gpt-3.5-turbo

references to recent time periods

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

 floors

-0.63

hod

-0.59

 halves

-0.59

66666666

-0.57

 desk

-0.56

NES

-0.56

onite

-0.55

ym

-0.55

 booth

-0.55

 Hunts

-0.55

POSITIVE LOGITS

 onwards

0.76

furt

0.74

hots

0.74

iannopoulos

0.73

iatus

0.73

pring

0.72

é¾įåĸļå£«

0.68

hips

0.68

cape

0.66

£ı

0.65

Activations Density 0.055%

phrases related to recent years

references to recent time periods

No Comments

No Known Activations