INDEX

Explanations

mentions of a person named "Ul" in various contexts

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

 Butterfly

-0.71

 Titanic

-0.70

 Turtles

-0.67

ãģĨ

-0.66

 Carbuncle

-0.63

xual

-0.63

 crib

-0.62

éĹĺ

-0.62

 Primal

-0.61

 Turtle

-0.61

POSITIVE LOGITS

terior

1.09

cers

0.97

gew

0.94

rican

0.91

leanor

0.90

lesiastical

0.90

umin

0.89

qus

0.89

aeper

0.87

usive

0.87

Activations Density 0.033%

mentions of a person named "Ul" in various contexts

No Comments

No Known Activations