INDEX

Explanations

words related to negative moral judgments or condemnations

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of being 'ominous' or 'ominousness.'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ħĭ

-0.72

 Leilan

-0.70

 distance

-0.67

 Sparrow

-0.64

 hint

-0.64

 Archangel

-0.63

 Higgins

-0.61

 mileage

-0.61

 Breaking

-0.61

¶ħ

-0.61

POSITIVE LOGITS

omin

1.50

ations

1.20

atory

1.14

itives

1.13

ational

1.08

etric

1.05

ices

1.01

omial

1.01

ous

1.01

itive

1.00

Activations Density 0.008%

words related to negative moral judgments or condemnations

terms related to the concept of being 'ominous' or 'ominousness.'

No Comments

No Known Activations