INDEX

Explanations

references to specific numerical statistics in sports such as points and percentages

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "from" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

quartered

-0.74

heit

-0.72

azeera

-0.72

ema

-0.71

ratulations

-0.70

merce

-0.70

bush

-0.69

idav

-0.68

igent

-0.67

irm

-0.65

POSITIVE LOGITS

 afar

1.40

 whence

0.93

 scratch

0.93

 scrimmage

0.83

 abroad

0.82

 inception

0.78

 inside

0.78

 anywhere

0.77

 atop

0.75

 across

0.73

Activations Density 0.150%

references to specific numerical statistics in sports such as points and percentages

occurrences of the word "from" in various contexts

No Comments

No Known Activations