INDEX

Explanations

the name "Luka" followed by either medical terms or other words

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or entities, particularly those with the name "uka."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

matically

-0.80

spin

-0.70

checks

-0.70

sc

-0.70

story

-0.69

izoph

-0.68

hip

-0.67

tem

-0.67

lines

-0.66

 ancest

-0.66

POSITIVE LOGITS

uka

1.00

Å¡

0.95

emia

0.93

iba

0.93

ichi

0.91

ovych

0.87

enko

0.86

ÅĤ

0.80

Ã±a

0.78

oji

0.77

Activations Density 0.010%

the name "Luka" followed by either medical terms or other words

references to specific individuals or entities, particularly those with the name "uka."

No Comments

No Known Activations