INDEX

Explanations

proper nouns related to the name "Kan" followed by a number

oai_token-act-pair · gpt-3.5-turbo

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

Ö¼

-0.93

é¾

-0.72

suit

-0.66

IBLE

-0.64

×Ļ×

-0.63

inately

-0.63

 Thumbnails

-0.61

bably

-0.61

 Aires

-0.61

çĭ

-0.60

POSITIVE LOGITS

ji

1.06

hof

1.06

awa

1.01

ako

0.98

amaru

0.91

mare

0.91

igan

0.90

ataka

0.89

ahi

0.89

ju

0.89

Activations Density 0.031%

proper nouns related to the name "Kan" followed by a number

No Comments

No Known Activations