INDEX

Explanations

phrases mentioning specific dates and events

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific character or symbol

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 farmland

-0.73

 polio

-0.71

 wartime

-0.70

 ammon

-0.70

 unwanted

-0.69

 resemblance

-0.69

 capacity

-0.69

 Mobil

-0.68

 mushroom

-0.68

 elusive

-0.68

POSITIVE LOGITS

efe

1.13

tis

1.13

ï¸ı

1.09

cause

1.09

she

1.08

nor

1.04

sn

1.02

ski

0.93

STEM

0.92

ometimes

0.92

Activations Density 0.180%

phrases mentioning specific dates and events

occurrences of a specific character or symbol

No Comments

No Known Activations