INDEX

Explanations

instances of the word 'many'

oai_token-act-pair · gpt-3.5-turbo

references to the quantity of things or people

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

ħĭ

-0.91

UAL

-0.74

acus

-0.70

Ħ¢

-0.68

TM

-0.67

ADVERTISEMENT

-0.67

works

-0.65

icer

-0.62

_-

-0.62

saf

-0.62

POSITIVE LOGITS

 times

1.02

 unanswered

0.88

ones

0.85

body

0.76

 dozen

0.73

 thousand

0.70

een

0.68

 hundred

0.68

 people

0.68

 facets

0.68

Activations Density 0.030%

instances of the word 'many'

references to the quantity of things or people

No Comments

No Known Activations