INDEX

Explanations

phrases indicating success or effectiveness

oai_token-act-pair · gpt-3.5-turbo

expressions indicating effectiveness or success

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

ategory

-0.75

agine

-0.69

htaking

-0.69

rush

-0.68

ilities

-0.67

guyen

-0.67

hyde

-0.66

avorite

-0.66

amera

-0.66

ilion

-0.65

POSITIVE LOGITS

 enough

1.30

enough

1.19

 Enough

0.91

bye

0.81

 behaved

0.80

baum

0.80

esley

0.77

spring

0.77

 suited

0.75

 vers

0.69

Activations Density 0.039%

phrases indicating success or effectiveness

expressions indicating effectiveness or success

No Comments

No Known Activations