INDEX

Explanations

specific company or organization names

oai_token-act-pair · gpt-3.5-turbo

names of organizations and companies

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

istics

-0.79

izations

-0.78

anza

-0.75

ruction

-0.72

ization

-0.71

eatures

-0.70

culosis

-0.67

orically

-0.66

Ñĭ

-0.65

eering

-0.65

POSITIVE LOGITS

ilon

0.85

tera

0.80

boro

0.77

acet

0.76

mos

0.76

hire

0.75

ulic

0.74

gpu

0.74

borough

0.74

ulla

0.73

Activations Density 0.032%

specific company or organization names

names of organizations and companies

No Comments

No Known Activations