INDEX

Explanations

mentions of infamous figures and political events

oai_token-act-pair · gpt-3.5-turbo

references to political figures and significant events

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

and

-0.64

and

-0.48

zona

-0.47

emia

-0.42

due

-0.42

 because

-0.42

ces

-0.42

itutional

-0.42

AND

-0.41

igent

-0.41

POSITIVE LOGITS

?ãĢį

0.73

âĢ

0.63

Instead

0.62

 instead

0.60

âĸł

0.59

SPONSORED

0.59

Elsewhere

0.59

Â·Â·

0.57

 sooner

0.56

æł

0.56

Activations Density 1.348%

mentions of infamous figures and political events

references to political figures and significant events

No Comments

No Known Activations