INDEX

Explanations

numbers in contexts related to specific statistics or quantities

oai_token-act-pair · gpt-3.5-turbo

numerical data and statistics

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

ult

-0.73

tre

-0.73

chat

-0.72

andr

-0.71

odon

-0.69

chwitz

-0.69

folk

-0.68

attery

-0.68

illet

-0.67

alam

-0.65

POSITIVE LOGITS

eenth

0.84

 cents

0.77

iffe

0.72

 spots

0.70

 percent

0.69

irlf

0.68

 clicks

0.67

 steps

0.65

 positives

0.64

 shoes

0.64

Activations Density 0.120%

numbers in contexts related to specific statistics or quantities

numerical data and statistics

No Comments

No Known Activations