INDEX

Explanations

mentions of the word "rab" or variations of it

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or entities associated with political discourse

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

éĹĺ

-0.69

 Tigers

-0.69

 Ceres

-0.67

 Resurrection

-0.66

 Occupations

-0.65

 Swordsman

-0.64

 appropriation

-0.64

 Spartan

-0.64

 Tsukuyomi

-0.63

 Spartans

-0.63

POSITIVE LOGITS

ble

1.00

inson

0.98

bing

0.96

iotics

0.93

ozo

0.89

anus

0.88

ody

0.87

orough

0.83

owitz

0.80

rab

0.79

Activations Density 0.016%

mentions of the word "rab" or variations of it

references to specific individuals or entities associated with political discourse

No Comments

No Known Activations