INDEX

Explanations

references to the word "Angel", potentially related to various contexts or entities

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Angel" and its variations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

yip

-0.75

merce

-0.67

rences

-0.67

Ĥª

-0.63

 independents

-0.63

ratulations

-0.61

 elig

-0.61

 disadvant

-0.60

 pressures

-0.58

rers

-0.57

POSITIVE LOGITS

enos

1.46

eno

1.23

ique

1.00

 Metatron

0.97

ica

0.96

angel

0.95

ion

0.95

ina

0.94

icals

0.94

opoulos

0.93

Activations Density 0.031%

references to the word "Angel", potentially related to various contexts or entities

mentions of the word "Angel" and its variations

No Comments

No Known Activations