INDEX

Explanations

proper nouns, particularly names

oai_token-act-pair · gpt-3.5-turbo

names and references related to specific individuals and factors connected to them

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

bre

-0.74

arov

-0.73

ĸļ

-0.73

igel

-0.70

ãĥĩãĤ£

-0.68

berra

-0.67

lies

-0.67

uyomi

-0.67

encers

-0.66

enza

-0.66

POSITIVE LOGITS

LR

1.12

LR

1.02

 Lank

0.75

 SECTION

0.74

SE

0.72

RAM

0.70

riott

0.69

andom

0.66

NX

0.66

outine

0.66

Activations Density 0.022%

proper nouns, particularly names

names and references related to specific individuals and factors connected to them

No Comments

No Known Activations