INDEX

Explanations

exact numbers preceded by the word "about"

oai_token-act-pair · gpt-3.5-turbo

phrases that provide approximate numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

NEY

-0.65

ters

-0.65

achus

-0.65

 Solution

-0.65

 Bomber

-0.64

OPA

-0.64

ses

-0.64

 Comed

-0.62

ande

-0.62

heses

-0.62

POSITIVE LOGITS

0.87

0.83

0.82

capacity

0.76

0.75

 Â½

0.74

 cubic

0.73

 half

0.73

0.72

Activations Density 0.088%

exact numbers preceded by the word "about"

phrases that provide approximate numerical values

No Comments

No Known Activations