INDEX

Explanations

O tokens

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "O" in various contexts

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

 commons

-0.70

 railing

-0.66

 downs

-0.60

 sharing

-0.60

 fodder

-0.59

 shared

-0.59

 stewards

-0.58

 scratching

-0.57

 chained

-0.56

 applicable

-0.56

POSITIVE LOGITS

3.63

Os

2.47

OX

1.83

OA

1.76

OE

1.71

1.62

HO

1.51

1.47

oS

1.45

OY

1.42

Activations Density 0.022%

O tokens

occurrences of the letter "O" in various contexts

No Comments

No Known Activations