INDEX

Explanations

numbers denoting quantities or statistics

oai_token-act-pair · gpt-3.5-turbo

numerical values indicating quantities or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

folk

-0.72

rium

-0.69

verages

-0.69

bear

-0.68

ment

-0.67

bryce

-0.67

Fram

-0.67

liest

-0.66

medi

-0.64

par

-0.64

POSITIVE LOGITS

 percent

1.04

%"

0.97

 consecutive

0.92

0.91

 hours

0.88

 thousand

0.87

0.85

 kilometers

0.82

 degrees

0.81

%.

0.81

Activations Density 0.126%

numbers denoting quantities or statistics

numerical values indicating quantities or statistics

No Comments

No Known Activations