INDEX

Explanations

phrases related to emphasizing a main idea or argument

oai_token-act-pair · gpt-3.5-turbo

statements centered around the concept of a "point" or central idea being made 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

uthor

-0.79

 Seym

-0.74

eco

-0.72

annis

-0.71

 destro

-0.70

reditary

-0.69

eatures

-0.69

ctic

-0.68

velt

-0.67

 notor

-0.67

POSITIVE LOGITS

points

0.88

point

0.86

lessly

0.86

iasis

0.86

lessness

0.81

 Reyes

0.80

posts

0.78

blank

0.76

manship

0.76

 point

0.75

Activations Density 0.039%

phrases related to emphasizing a main idea or argument

statements centered around the concept of a "point" or central idea being made in various contexts

No Comments

No Known Activations