INDEX

Explanations

phrases related to public statements and apologies

oai_token-act-pair · gpt-3.5-turbo

phrases indicating responsibility or accountability in statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.83

arag

-0.67

ready

-0.67

 clout

-0.66

kefeller

-0.65

 doom

-0.62

yip

-0.62

phabet

-0.62

abal

-0.61

hack

-0.61

POSITIVE LOGITS

Statement

0.98

 regrett

0.94

Our

0.93

 apologize

0.92

Our

0.88

Upon

0.86

 apologies

0.85

 hereby

0.85

Due

0.84

 unres

0.83

Activations Density 0.506%

phrases related to public statements and apologies

phrases indicating responsibility or accountability in statements

No Comments

No Known Activations