INDEX

Explanations

mentions of something new or a change

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "new"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Previously

-0.77

 Originally

-0.76

umenthal

-0.73

 previously

-0.72

ertodd

-0.72

cius

-0.72

ouse

-0.69

lua

-0.69

olen

-0.68

Zip

-0.66

POSITIVE LOGITS

bies

1.34

bie

1.30

 arrivals

1.22

 millennium

1.13

 hires

0.97

 incarnation

0.96

 entrants

0.95

 generation

0.92

foundland

0.88

 additions

0.85

Activations Density 0.097%

mentions of something new or a change

occurrences of the word "new"

No Comments

No Known Activations