INDEX

Explanations

commands or instructions starting with "Don't."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "don't."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ħĭ

-0.86

£ı

-0.82

ĪĴ

-0.77

 behavi

-0.76

 Reloaded

-0.75

Reviewer

-0.75

 ingred

-0.73

 shorth

-0.70

milo

-0.68

 ancest

-0.67

POSITIVE LOGITS

cha

0.89

 bother

0.87

osaurs

0.84

ables

0.83

rave

0.82

ween

0.81

ardless

0.81

ional

0.81

reprene

0.80

urb

0.79

Activations Density 0.055%

commands or instructions starting with "Don't."

occurrences of the word "don't."

No Comments

No Known Activations