INDEX

Explanations

lists of items or topics

oai_token-act-pair · gpt-3.5-turbo

references to various items or categories represented by "etc."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Lords

-0.78

inka

-0.75

adows

-0.70

wives

-0.61

 shadow

-0.60

 Broad

-0.60

Gos

-0.59

-0.58

lled

-0.57

FX

-0.57

POSITIVE LOGITS

etc

1.17

eter

1.08

.?

0.90

etc

0.90

.,

0.89

gow

0.89

arently

0.83

ignt

0.83

ãģĨ

0.80

uits

0.79

Activations Density 0.017%

lists of items or topics

references to various items or categories represented by "etc."

No Comments

No Known Activations