INDEX

Explanations

words associated with creepiness or unsettling content

oai_token-act-pair · gpt-3.5-turbo

the presence of the letter 'y' in various contexts

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

 Examiner

-0.84

insula

-0.79

raltar

-0.76

ãĥ¯

-0.70

 "$:/

-0.70

 PowerPoint

-0.70

 ensu

-0.67

vou

-0.66

ĸļ

-0.64

 FEMA

-0.64

POSITIVE LOGITS

ield

1.09

ielding

1.04

ikes

0.95

ank

0.91

aky

0.90

Å«

0.89

aku

0.85

unk

0.84

asha

0.84

kee

0.82

Activations Density 0.044%

words associated with creepiness or unsettling content

the presence of the letter 'y' in various contexts

No Comments

No Known Activations