INDEX

Explanations

details that are formatted as bullet points or subheadings within a longer text

oai_token-act-pair · gpt-3.5-turbo

bullet points or lists in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

udic

-0.81

othal

-0.79

erer

-0.74

aults

-0.67

ierre

-0.67

uve

-0.66

ERC

-0.63

aughter

-0.62

olyn

-0.61

enthal

-0.60

POSITIVE LOGITS

Â·Â·

1.25

âĢ¢âĢ¢

0.82

¼

0.76

¾

0.72

 Joined

0.71

âĢ¢âĢ¢âĢ¢âĢ¢

0.71

ting

0.70

thia

0.70

µ

0.68

lat

0.67

Activations Density 0.014%

details that are formatted as bullet points or subheadings within a longer text

bullet points or lists in the text

No Comments

No Known Activations