INDEX

Explanations

phrases that mention a quantification or comparison of a group of people or things

oai_token-act-pair · gpt-3.5-turbo

references to large quantities or populations 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

imir

-0.78

aunder

-0.66

 Dull

-0.61

Rav

-0.61

 Gutenberg

-0.61

 countdown

-0.61

 glim

-0.60

Enh

-0.60

adobe

-0.58

 Aviv

-0.58

POSITIVE LOGITS

(>

1.04

(<

0.74

(~

0.66

IU

0.65

 majorities

0.64

ONES

0.63

 populous

0.63

agos

0.63

atoes

0.63

ortment

0.60

Activations Density 0.163%

phrases that mention a quantification or comparison of a group of people or things

references to large quantities or populations in various contexts

No Comments

No Known Activations