INDEX

Explanations

names of locations or people

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Burns" and associated names, likely identifying locations or individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

eanor

-0.77

rium

-0.71

awaru

-0.68

eva

-0.68

universal

-0.67

apped

-0.66

ilial

-0.65

ascal

-0.65

remlin

-0.65

pai

-0.63

POSITIVE LOGITS

 Burns

1.37

 Burn

0.93

lett

0.87

Burn

0.83

lee

0.77

leys

0.76

worth

0.74

 Timbers

0.73

ford

0.73

 burns

0.71

Activations Density 0.007%

names of locations or people

mentions of the word "Burns" and associated names, likely identifying locations or individuals

No Comments

No Known Activations