INDEX

Explanations

mentions of the word "Lac."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Lac" and its variants in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ITNESS

-0.67

FTWARE

-0.65

ãĥīãĥ©

-0.65

fulness

-0.64

izons

-0.62

icals

-0.61

 Shinra

-0.61

à¥

-0.59

 Hancock

-0.59

ãĥĦ

-0.59

POSITIVE LOGITS

rosse

1.22

Lac

1.17

lac

0.93

iple

0.85

quer

0.84

kered

0.82

 mathemat

0.81

 municip

0.79

orem

0.78

rylic

0.76

Activations Density 0.006%

mentions of the word "Lac."

occurrences of the term "Lac" and its variants in various contexts

No Comments

No Known Activations