INDEX

Explanations

words related to different variations of the term "Kiw" and "William"

oai_token-act-pair · gpt-3.5-turbo

occurrences of specific identifiers or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

vier

-0.73

izing

-0.70

âĸ¬

-0.70

 Surgery

-0.68

 surgery

-0.67

ized

-0.65

heimer

-0.60

izers

-0.59

isers

-0.59

izes

-0.59

POSITIVE LOGITS

awei

0.91

iard

0.87

arta

0.78

tis

0.77

iaz

0.76

entric

0.75

eas

0.75

ao

0.75

alos

0.75

adr

0.74

Activations Density 0.035%

words related to different variations of the term "Kiw" and "William"

occurrences of specific identifiers or entities

No Comments

No Known Activations