INDEX

Explanations

instances of the word "leap" where the activation value is high

oai_token-act-pair · gpt-3.5-turbo

instances of the word "leap" 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 @ 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

essee

-1.08

gel

-0.80

Interstitial

-0.78

ividually

-0.70

pmwiki

-0.67

ĻĤ

-0.66

liction

-0.66

gew

-0.64

matter

-0.63

leanor

-0.62

POSITIVE LOGITS

frog

1.14

 leaps

1.02

 leap

0.79

olicy

0.79

rack

0.76

 Leap

0.74

rers

0.73

 Rivals

0.69

fruit

0.69

 forward

0.68

Activations Density 0.017%

instances of the word "leap" where the activation value is high

instances of the word "leap" in various contexts

No Comments

No Known Activations