INDEX

Explanations

places or entities that come in pairs or groups of two

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or groups that are quantified using the word "two."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

taboola

-0.86

ople

-0.66

 EVENTS

-0.65

SPONSORED

-0.61

humane

-0.61

Ĥİ

-0.58

activity

-0.57

à¤

-0.55

Fre

-0.55

nance

-0.55

POSITIVE LOGITS

 apiece

1.22

 totaling

1.02

paces

0.92

hips

0.91

 respectively

0.89

 simultaneously

0.89

 consecut

0.86

 separated

0.83

 collided

0.80

poons

0.80

Activations Density 0.289%

places or entities that come in pairs or groups of two

references to specific entities or groups that are quantified using the word "two."

No Comments

No Known Activations