INDEX

Explanations

sections of text referencing specific numerical data or points within a larger document

oai_token-act-pair · gpt-3.5-turbo

sentence-ending punctuation marks

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

ãĥ¼ãĥĨ

-0.84

ãĥ¼ãĥĨãĤ£

-0.75

 seams

-0.70

ãĥ¼ãĥ³

-0.66

 Franch

-0.65

 plantation

-0.64

 patri

-0.63

ikuman

-0.62

iquette

-0.62

çĲ

-0.61

POSITIVE LOGITS

...]

1.49

âĢ¦]

1.46

Pg

1.27

?]

1.08

emphasis

1.07

!]

0.99

Footnote

0.98

REDACTED

0.94

sic

0.93

:]

0.93

Activations Density 0.034%

sections of text referencing specific numerical data or points within a larger document

sentence-ending punctuation marks

No Comments

No Known Activations