INDEX

Explanations

mentions of the term "IGN" with varying degrees of detail and context

oai_token-act-pair · gpt-3.5-turbo

mentions of the organization or brand "IGN"

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

 Maid

-0.72

zynski

-0.69

 Sask

-0.67

bian

-0.67

 separatists

-0.63

foot

-0.62

 Nurs

-0.62

 Kenyan

-0.62

 wings

-0.62

 Finnish

-0.62

POSITIVE LOGITS

IGN

1.35

IGN

0.98

VERTISEMENT

0.93

ution

0.87

arious

0.85

TY

0.85

ixel

0.85

ORED

0.85

ito

0.83

uci

0.83

Activations Density 0.008%

mentions of the term "IGN" with varying degrees of detail and context

mentions of the organization or brand "IGN"

No Comments

No Known Activations