INDEX

Explanations

words related to the Jewish faith and culture

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Sep" used as a date reference

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

culosis

-0.98

 tuberculosis

-0.83

Reviewer

-0.76

 Archdemon

-0.73

ãĥīãĥ©ãĤ´ãĥ³

-0.71

hyde

-0.65

anza

-0.65

ngth

-0.65

 Jackets

-0.63

 modem

-0.63

POSITIVE LOGITS

arate

1.15

aration

1.10

ay

0.95

aret

0.92

ul

0.85

rees

0.84

ayed

0.82

otiation

0.81

arat

0.80

Ãº

0.80

Activations Density 0.007%

words related to the Jewish faith and culture

occurrences of the term "Sep" used as a date reference

No Comments

No Known Activations