INDEX

Explanations

references to occurrences in different locations or environments

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in" within various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 accordingly

-0.79

something

-0.76

 likewise

-0.74

 instead

-0.72

 even

-0.71

almost

-0.70

even

-0.70

 almost

-0.70

 whenever

-0.70

POSITIVE LOGITS

 sexes

1.09

 genders

0.87

 sender

0.78

 physical

0.75

 academia

0.73

BuyableInstoreAndOnline

0.73

 literal

0.69

 textual

0.68

 verbal

0.68

 hardware

0.67

Activations Density 0.169%

references to occurrences in different locations or environments

occurrences of the word "in" within various contexts

No Comments

No Known Activations