INDEX

Explanations

supporter-related terms and phrases

oai_token-act-pair · gpt-3.5-turbo

references to political supporters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

aber

-0.71

iencies

-0.69

posed

-0.68

ofi

-0.68

posure

-0.66

Mars

-0.66

Pain

-0.65

 mountainous

-0.64

ospital

-0.63

apes

-0.63

POSITIVE LOGITS

 supporter

1.27

 supporters

1.05

Supporters

1.04

 Supporters

1.00

hesis

0.93

="#

0.90

 backer

0.87

hetical

0.86

isSpecialOrderable

0.81

 backers

0.80

Activations Density 0.005%

supporter-related terms and phrases

references to political supporters

No Comments

No Known Activations