INDEX

Explanations

locations related to a specific geographical area

oai_token-act-pair · gpt-3.5-turbo

punctuation, specifically periods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

plet

-0.82

 mosqu

-0.80

 microbiome

-0.80

 tack

-0.80

 upgr

-0.77

 pace

-0.76

urance

-0.76

 crunch

-0.75

 tackling

-0.74

 welcome

-0.72

POSITIVE LOGITS

 Interestingly

1.52

 However

1.52

 Although

1.44

 Occasionally

1.44

 Additionally

1.40

 Though

1.38

 According

1.38

 Originally

1.37

 Ironically

1.36

 Notably

1.36

Activations Density 0.309%

locations related to a specific geographical area

punctuation, specifically periods

No Comments

No Known Activations