INDEX

Explanations

proper nouns related to various organizations or individuals

oai_token-act-pair · gpt-3.5-turbo

specific characters or names used in titles and proper nouns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 milit

-0.71

¥ŀ

-0.70

 partition

-0.67

 mould

-0.66

 segregated

-0.64

 propag

-0.64

 safegu

-0.64

 cannibal

-0.64

 orphan

-0.63

spawn

-0.62

POSITIVE LOGITS

oglu

0.79

anche

0.79

ulo

0.78

olor

0.78

idis

0.78

 guiActiveUnfocused

0.76

ians

0.76

=#

0.76

imum

0.75

ogl

0.75

Activations Density 0.355%

proper nouns related to various organizations or individuals

specific characters or names used in titles and proper nouns

No Comments

No Known Activations