INDEX

Explanations

mentions of the word "Hope"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Hope."

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

atable

-0.80

otomy

-0.80

astical

-0.78

ickr

-0.76

aton

-0.73

azes

-0.73

ographically

-0.72

astically

-0.72

chnology

-0.71

ographers

-0.70

POSITIVE LOGITS

ful

1.20

lyn

0.87

 Hills

0.85

 Intervention

0.84

fulness

0.83

shire

0.81

 Valley

0.76

keeper

0.76

 Falls

0.76

 Talks

0.75

Activations Density 0.033%

mentions of the word "Hope"

instances of the word "Hope."

No Comments

No Known Activations