INDEX

Explanations

phrases containing the word 'banned'

oai_token-act-pair · gpt-3.5-turbo

words related to restrictions or bans

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

du

-0.67

 present

-0.65

 guard

-0.64

fi

-0.63

 Knight

-0.63

XII

-0.62

 symptom

-0.61

 injury

-0.60

 stab

-0.60

Cor

-0.59

POSITIVE LOGITS

anned

4.45

anners

1.41

rawled

1.12

ans

1.11

aned

1.07

attered

1.03

illed

1.01

aced

0.99

aled

0.97

outed

0.96

Activations Density 0.011%

phrases containing the word 'banned'

words related to restrictions or bans

No Comments

No Known Activations