INDEX

Explanations

phrases related to putting things in perspective or into place

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate context or situational placement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ENT

-0.71

zan

-0.69

enance

-0.69

ences

-0.69

ľ

-0.68

LEY

-0.67

ENTS

-0.67

ents

-0.65

bey

-0.63

uren

-0.63

POSITIVE LOGITS

 jeopardy

1.17

 perspective

0.99

 place

0.98

 context

0.92

 quotation

0.83

 harms

0.81

 motion

0.79

 orbit

0.75

 charge

0.74

 front

0.74

Activations Density 0.088%

phrases related to putting things in perspective or into place

phrases that indicate context or situational placement

No Comments

No Known Activations