INDEX

Explanations

words related to lists or items

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 's'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fuss

-0.74

£ı

-0.70

 Leilan

-0.68

ufact

-0.67

electric

-0.67

IOR

-0.65

icio

-0.65

RFC

-0.64

 actionGroup

-0.64

ilater

-0.64

POSITIVE LOGITS

kaya

1.17

ources

1.06

ourcing

1.05

ourced

1.02

linger

1.01

wered

0.90

atisf

0.89

hips

0.88

ouls

0.87

leeve

0.86

Activations Density 0.237%

words related to lists or items

occurrences of the letter 's'

No Comments

No Known Activations