INDEX

Explanations

phrases related to complex relationships between entities or concepts

oai_token-act-pair · gpt-3.5-turbo

references to relationships between different entities

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OGR

-0.76

ONSORED

-0.72

vous

-0.72

Lev

-0.70

bed

-0.69

br

-0.68

der

-0.67

ski

-0.67

herty

-0.66

rats

-0.66

POSITIVE LOGITS

between

0.95

 sexes

0.81

 between

0.81

 Between

0.78

 halves

0.68

 genders

0.67

 friendships

0.61

ween

0.61

 rounds

0.61

 Kore

0.60

Activations Density 0.049%

phrases related to complex relationships between entities or concepts

references to relationships between different entities

No Comments

No Known Activations