INDEX

Explanations

phrases related to public speaking or presentations

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly exclamation points and question marks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Darth

-0.57

 grav

-0.53

-0.51

 Lump

-0.51

Alert

-0.49

 public

-0.48

 road

-0.47

 craft

-0.47

 Tesla

-0.47

Ant

-0.47

POSITIVE LOGITS

!,

2.91

?,

2.60

!.

2.09

/,

2.07

$,

1.82

*,

1.68

!",

1.68

Â®,

1.58

(),

1.56

+,

1.56

Activations Density 0.023%

phrases related to public speaking or presentations

punctuation marks, particularly exclamation points and question marks

No Comments

No Known Activations