INDEX

Explanations

language related to categorization and belonging, such as membership, relatedness, race, and elites

oai_token-act-pair · gpt-3.5-turbo

references to ethnic or cultural group membership

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

effic

-0.73

displayText

-0.68

charges

-0.68

 unanswered

-0.66

 unfolds

-0.64

Results

-0.62

efficiency

-0.62

Everything

-0.61

hitting

-0.61

 fumble

-0.61

POSITIVE LOGITS

 hapl

1.04

 caste

1.00

 ethnic

0.99

groups

0.98

 category

0.98

 sects

0.96

 minority

0.96

 subclass

0.95

 sect

0.95

 genus

0.94

Activations Density 0.289%

language related to categorization and belonging, such as membership, relatedness, race, and elites

references to ethnic or cultural group membership

No Comments

No Known Activations