INDEX

Explanations

the word fragments that end with "ka."

oai_token-act-pair · gpt-3.5-turbo

repeated names or terms ending with "ka."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lessly

-0.79

Ö¼

-0.75

orship

-0.75

á½

-0.72

reads

-0.69

advertisement

-0.67

aments

-0.64

service

-0.63

down

-0.62

 Murdoch

-0.61

POSITIVE LOGITS

ichi

1.12

osta

0.99

esthesia

0.92

iba

0.87

esthetic

0.86

uthor

0.83

jac

0.82

oru

0.80

TeX

0.78

vel

0.76

Activations Density 0.009%

the word fragments that end with "ka."

repeated names or terms ending with "ka."

No Comments

No Known Activations