INDEX

Explanations

words related to personal names, potentially of people or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns and names often related to individuals and organizations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

é¾įå¥ĳå£«

-0.74

»Ĵ

-0.72

ãģ¦

-0.70

 cellul

-0.69

dayName

-0.65

 bluff

-0.65

 stewards

-0.63

ĨĴ

-0.61

Maria

-0.60

WARN

-0.60

POSITIVE LOGITS

andro

0.77

igne

0.74

ocene

0.73

orce

0.73

oscopic

0.70

oche

0.70

tics

0.69

velt

0.69

coe

0.69

ograp

0.68

Activations Density 0.079%

words related to personal names, potentially of people or places

proper nouns and names often related to individuals and organizations

No Comments

No Known Activations