INDEX

Explanations

phrases related to cultural references, names, and titles

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the term "ura" 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 @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

rodu

-0.96

mosp

-0.80

raltar

-0.80

oration

-0.79

sonian

-0.76

tons

-0.76

ablishment

-0.76

herer

-0.75

oyer

-0.73

rate

-0.72

POSITIVE LOGITS

Downloadha

0.75

pmwiki

0.75

 Mazda

0.73

BIP

0.68

 este

0.67

Zar

0.67

Ã°

0.67

qqa

0.66

BILITY

0.65

igi

0.64

Activations Density 0.041%

phrases related to cultural references, names, and titles

repeated mentions of the term "ura" in various contexts

No Comments

No Known Activations