INDEX

Explanations

specific mention of the name "Rad" with varying intensities of activations

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Rad" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Ń·

-0.78

redits

-0.75

terday

-0.75

ORTS

-0.75

pload

-0.69

urities

-0.68

asper

-0.67

FAULT

-0.66

holding

-0.63

âĢ¢âĢ¢

-0.63

POSITIVE LOGITS

iance

1.11

isson

1.03

ient

1.01

iation

0.98

iosyn

0.97

iated

0.96

iop

0.95

isco

0.93

iate

0.93

iations

0.92

Activations Density 0.012%

specific mention of the name "Rad" with varying intensities of activations

mentions of the term "Rad" in various contexts

No Comments

No Known Activations