INDEX

Explanations

mentions of "scar" or variations of the word

oai_token-act-pair · gpt-3.5-turbo

terms related to scar tissue and its implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ITNESS

-0.67

 Fathers

-0.60

ãģ®å®

-0.59

ECTION

-0.59

onyms

-0.58

 OTHER

-0.58

ERAL

-0.57

£ı

-0.56

vironment

-0.55

SOURCE

-0.55

POSITIVE LOGITS

scar

1.22

iform

0.86

vine

0.83

rio

0.83

acho

0.82

gio

0.79

ques

0.79

otti

0.78

ousse

0.78

ring

0.78

Activations Density 0.005%

mentions of "scar" or variations of the word

terms related to scar tissue and its implications

No Comments

No Known Activations