INDEX

Explanations

patterns related to mathematical or programming terms and symbols

oai_token-act-pair · gpt-3.5-turbo

conjunctions and transitional phrases that connect thoughts or ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

izzle

-0.74

uits

-0.74

ĪĴ

-0.72

ounding

-0.72

ĸļ

-0.69

helm

-0.67

ãĤ¡

-0.67

ens

-0.66

uca

-0.66

abouts

-0.65

POSITIVE LOGITS

 Conversely

0.85

 whereas

0.80

 Whereas

0.78

 others

0.69

 Britann

0.66

Medium

0.66

 Others

0.66

 ours

0.65

 Monarch

0.64

 Anon

0.62

Activations Density 0.449%

patterns related to mathematical or programming terms and symbols

conjunctions and transitional phrases that connect thoughts or ideas

No Comments

No Known Activations