INDEX

Explanations

patterns related to Hebrew characters and words

oai_token-act-pair · gpt-3.5-turbo

textual representations of non-English characters or scripts

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

lyak

-0.78

Cly

-0.77

upiter

-0.76

estern

-0.74

annis

-0.74

iko

-0.73

arching

-0.72

tten

-0.69

psey

-0.67

onge

-0.67

POSITIVE LOGITS

Ù

1.74

ÙĨ

1.66

Ùĩ

1.66

Ø§

1.62

Ø¯

1.57

Ùħ

1.55

Ø§Ø

1.54

Ø

1.52

ÙĬ

1.50

Øª

1.49

Activations Density 0.004%

patterns related to Hebrew characters and words

textual representations of non-English characters or scripts

No Comments

No Known Activations