INDEX

Explanations

text that suggests the reader to take certain actions or make certain decisions

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases or expressions indicating a transition in discussion or action

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

 decomp

-0.66

 cyan

-0.66

 shack

-0.64

enegger

-0.63

 unmarked

-0.63

 lifestyle

-0.63

 guarding

-0.63

 minim

-0.62

FAT

-0.61

Downloadha

-0.60

POSITIVE LOGITS

¹

0.97

º

0.97

ı

0.93

ISIS

0.90

į

0.84

âĢķ

0.81

½

0.80

Ľ

0.80

Ī

0.80

_>

0.80

Activations Density 0.507%

text that suggests the reader to take certain actions or make certain decisions

repetitive phrases or expressions indicating a transition in discussion or action

No Comments

No Known Activations