INDEX

Explanations

names or terms mentioned as the subject of discussion or investigation

oai_token-act-pair · gpt-3.5-turbo

references to specific inquiries or topics being discussed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

azon

-0.82

rid

-0.76

atever

-0.74

ahime

-0.73

lez

-0.72

 millenn

-0.72

 newcom

-0.70

ERAL

-0.69

urate

-0.69

inders

-0.68

POSITIVE LOGITS

*/(

0.83

:(

0.67

 belonged

0.65

hess

0.63

oux

0.59

 belong

0.58

âĸĵ

0.57

 belongs

0.56

abwe

0.55

ube

0.54

Activations Density 0.103%

names or terms mentioned as the subject of discussion or investigation

references to specific inquiries or topics being discussed

No Comments

No Known Activations