INDEX

Explanations

instances of the word "Another" followed by a continuation or list of items

oai_token-act-pair · gpt-3.5-turbo

phrases that introduce new points or arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

hips

-0.82

ouls

-0.82

onies

-0.72

ivas

-0.70

olas

-0.69

alties

-0.68

obar

-0.67

riages

-0.67

present

-0.63

amy

-0.62

POSITIVE LOGITS

worldly

0.94

 aspect

0.94

 notable

0.86

 drawback

0.86

 example

0.86

 pecul

0.85

 notch

0.84

 factor

0.82

 thing

0.81

 unnamed

0.80

Activations Density 0.030%

instances of the word "Another" followed by a continuation or list of items

phrases that introduce new points or arguments

No Comments

No Known Activations