INDEX

Explanations

numerical data such as quantities or statistics

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "number" followed by numerical values or statistics

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

lace

-0.68

dit

-0.64

 germ

-0.64

romy

-0.61

din

-0.57

hem

-0.57

iolet

-0.56

INTON

-0.56

 backdrop

-0.55

rament

-0.55

POSITIVE LOGITS

of

0.93

 crunch

0.70

 dropped

0.68

 dwind

0.68

Of

0.63

enance

0.63

aurus

0.62

 drops

0.62

lessness

0.62

 number

0.61

Activations Density 0.056%

numerical data such as quantities or statistics

occurrences of the word "number" followed by numerical values or statistics

No Comments

No Known Activations