INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, especially names associated with people and places

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

 breath

-0.72

 Redux

-0.62

 electronically

-0.62

Constructed

-0.61

 rendition

-0.61

 centrif

-0.61

Madison

-0.61

idency

-0.61

 prevailing

-0.61

 speech

-0.61

POSITIVE LOGITS

atures

1.19

er

0.97

atical

0.94

owship

0.93

boat

0.93

ichick

0.93

atural

0.92

ular

0.90

atin

0.88

atics

0.88

Activations Density 0.043%

names of individuals

proper nouns, especially names associated with people and places

No Comments

No Known Activations