INDEX

Explanations

instances where the word "walls" is mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

references to walls and their conditions or features

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

inal

-0.65

onement

-0.65

atorium

-0.64

avez

-0.63

ya

-0.63

arin

-0.63

lich

-0.63

capital

-0.62

INA

-0.62

sb

-0.62

POSITIVE LOGITS

 walls

1.20

mith

1.08

creen

1.05

 Walls

0.98

papers

0.86

matter

0.82

paper

0.81

 wall

0.78

cape

0.77

 wallpaper

0.76

Activations Density 0.032%

instances where the word "walls" is mentioned in the text

references to walls and their conditions or features

No Comments

No Known Activations