INDEX

Explanations

codes or programming related terms

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Notice" and related instructions

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

 outwe

-0.83

 paran

-0.81

 Carlton

-0.73

awa

-0.73

 arri

-0.69

 puberty

-0.68

 assail

-0.68

 ambassadors

-0.68

 Crab

-0.67

 unexplained

-0.66

POSITIVE LOGITS

Output

1.01

PUT

0.95

Example

0.94

Usage

0.90

Frames

0.88

RAW

0.88

Console

0.87

Gener

0.86

{{

0.86

Returns

0.85

Activations Density 0.132%

codes or programming related terms

instances of the word "Notice" and related instructions

No Comments

No Known Activations