INDEX

Explanations

phrases related to traumatic or challenging experiences

oai_token-act-pair · gpt-3.5-turbo

words related to affirmation or confirmation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 è£ıç

-0.89

REDACTED

-0.73

Spoiler

-0.69

 Lights

-0.68

 Rasm

-0.67

 spared

-0.67

å¥

-0.66

 pandemonium

-0.65

 Tribunal

-0.64

NetMessage

-0.64

POSITIVE LOGITS

atively

1.17

irm

1.12

ative

1.07

ament

1.01

atives

0.95

aton

0.93

ware

0.86

irms

0.85

anyahu

0.85

atory

0.85

Activations Density 0.026%

phrases related to traumatic or challenging experiences

words related to affirmation or confirmation

No Comments

No Known Activations