INDEX

Explanations

words related to modifications or versions of something

oai_token-act-pair · gpt-3.5-turbo

references to "Mod" and "modular" concepts 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ģĸ

-0.83

éĽ

-0.73

ï¸

-0.73

 Lauder

-0.72

izabeth

-0.69

OHN

-0.68

Ã§a

-0.68

WAYS

-0.67

 Oaks

-0.67

 Mata

-0.66

POSITIVE LOGITS

icum

1.22

ding

1.16

ifiable

1.14

ifiers

1.14

ifications

1.13

ules

1.07

ifying

1.06

elled

1.05

ifies

1.04

ifier

1.04

Activations Density 0.009%

words related to modifications or versions of something

references to "Mod" and "modular" concepts in various contexts

No Comments

No Known Activations