INDEX

Explanations

references to a specific person named Woods

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual named Woods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ince

-0.74

 decl

-0.69

 registration

-0.67

 fare

-0.67

 transit

-0.63

 epile

-0.62

 transporter

-0.61

 droid

-0.61

ado

-0.60

 seizure

-0.60

POSITIVE LOGITS

 Woods

4.18

Wood

1.93

woods

1.58

 Wood

1.52

 woods

1.51

 Waters

1.37

 Oaks

1.36

 Meadows

1.25

 Hills

1.16

 Wilderness

1.16

Activations Density 0.017%

references to a specific person named Woods

references to a specific individual named Woods

No Comments

No Known Activations