INDEX

Explanations

country names or other geopolitical entities in conjunction with specific organizations, newspapers, or currencies

oai_token-act-pair · gpt-3.5-turbo

the word "and" across various contexts

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

elle

-0.76

IGHTS

-0.74

rimination

-0.74

Both

-0.71

Based

-0.70

iciary

-0.70

istical

-0.69

ogy

-0.69

ugi

-0.69

wic

-0.68

POSITIVE LOGITS

 others

0.83

 even

0.77

 perhaps

0.77

etc

0.76

 hence

0.75

 possibly

0.74

 assorted

0.73

 perennial

0.69

chard

0.69

 elsewhere

0.69

Activations Density 0.223%

country names or other geopolitical entities in conjunction with specific organizations, newspapers, or currencies

the word "and" across various contexts

No Comments

No Known Activations