INDEX

Explanations

references to individuals holding high-ranking political positions such as "premier."

oai_token-act-pair · gpt-3.5-turbo

the word "premier" as it relates to leaders or top entities 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 @ 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

arettes

-0.85

akia

-0.81

outs

-0.80

alle

-0.78

gers

-0.77

ename

-0.75

antry

-0.73

aped

-0.71

OUT

-0.71

IGHTS

-0.70

POSITIVE LOGITS

ois

0.97

iary

0.76

 constitu

0.75

ing

0.73

 contender

0.72

 Provincial

0.70

ially

0.70

ICS

0.70

 ministerial

0.69

 Kathleen

0.68

Activations Density 0.010%

references to individuals holding high-ranking political positions such as "premier."

the word "premier" as it relates to leaders or top entities in various contexts

No Comments

No Known Activations