INDEX

Explanations

phrases related to pursuing or targeting someone or something

oai_token-act-pair · gpt-3.5-turbo

the phrase "going after."

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

Released

-0.76

Chart

-0.73

uten

-0.73

aez

-0.72

Near

-0.69

Catalog

-0.69

Thirty

-0.69

UTION

-0.68

ï¸

-0.67

ailand

-0.66

POSITIVE LOGITS

market

0.85

wards

0.74

 tweaking

0.70

 whistleblowers

0.70

 installing

0.69

 delivering

0.68

 controlling

0.68

 inspecting

0.68

 deleting

0.67

 supper

0.67

Activations Density 0.036%

phrases related to pursuing or targeting someone or something

the phrase "going after."

No Comments

No Known Activations