INDEX

Explanations

words related to specific names, possibly of individuals or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns associated with specific locations or entities

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

stra

-0.76

sta

-0.75

kie

-0.74

fulness

-0.73

ful

-0.73

kies

-0.71

choes

-0.68

onen

-0.67

ãĤº

-0.67

fully

-0.66

POSITIVE LOGITS

 Mamm

0.87

è¦ļéĨĴ

0.73

 Whale

0.72

acent

0.72

 predec

0.71

axter

0.71

 Hamb

0.71

 charism

0.71

arella

0.68

odied

0.67

Activations Density 0.025%

words related to specific names, possibly of individuals or places

proper nouns associated with specific locations or entities

No Comments

No Known Activations