INDEX

Explanations

phrases related to unfavorable or embarrassing situations and events

oai_token-act-pair · gpt-3.5-turbo

references to critical situations or significant consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Modes

-0.72

 modes

-0.68

udicrous

-0.68

Scroll

-0.64

 cones

-0.63

renheit

-0.59

ilib

-0.59

atives

-0.59

 Balls

-0.57

 regress

-0.56

POSITIVE LOGITS

or

1.05

someone

1.00

 spouse

0.96

Someone

0.91

 maybe

0.90

maybe

0.88

 someone

0.86

 Someone

0.86

perhaps

0.83

 somebody

0.82

Activations Density 1.005%

phrases related to unfavorable or embarrassing situations and events

references to critical situations or significant consequences

No Comments

No Known Activations