INDEX

Explanations

terms related to concentration camps and focus on specific phrases related to them

oai_token-act-pair · gpt-3.5-turbo

references to concentration camps

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ĪĴ

-0.80

aired

-0.75

facing

-0.71

 Logged

-0.71

mic

-0.71

posted

-0.70

 ambassadors

-0.70

\\\\\\\\

-0.70

named

-0.68

anne

-0.67

POSITIVE LOGITS

 concentration

0.98

 Concent

0.80

itated

0.80

emetery

0.76

Pru

0.76

itate

0.75

 concentrated

0.74

itating

0.73

arnaev

0.72

inct

0.72

Activations Density 0.051%

terms related to concentration camps and focus on specific phrases related to them

references to concentration camps

No Comments

No Known Activations