INDEX

Explanations

mentions of residents in various locations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "resident"

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

anche

-0.98

lda

-0.84

enic

-0.83

matically

-0.81

anwhile

-0.81

anson

-0.77

arrell

-0.77

gettable

-0.76

Bone

-0.76

reme

-0.76

POSITIVE LOGITS

hips

0.80

 resident

0.78

iary

0.76

 Evil

0.75

ial

0.73

ially

0.73

iate

0.69

 Resident

0.68

iaries

0.67

resident

0.66

Activations Density 0.021%

mentions of residents in various locations

instances of the word "resident"

No Comments

No Known Activations