INDEX

Explanations

words related to structural elements or organization

oai_token-act-pair · gpt-3.5-turbo

references to structural concepts or terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

hood

-0.71

cale

-0.66

 Leone

-0.65

cli

-0.61

 clutch

-0.61

fly

-0.60

ously

-0.59

 overboard

-0.59

 Clarkson

-0.59

Yel

-0.59

POSITIVE LOGITS

urally

1.54

ured

1.31

ural

1.21

uration

1.12

uring

1.07

ures

1.05

urer

1.03

atile

1.03

untarily

1.00

rait

0.96

Activations Density 0.017%

words related to structural elements or organization

references to structural concepts or terminology

No Comments

No Known Activations