INDEX

Explanations

keywords or phrases related to physical keys

oai_token-act-pair · gpt-3.5-turbo

mentions of "keys" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ATIONAL

-0.78

ews

-0.72

 Colleges

-0.66

 Wolver

-0.65

AUT

-0.65

ESCO

-0.65

nn

-0.65

asca

-0.63

OUS

-0.62

 Constantin

-0.62

POSITIVE LOGITS

 keys

1.34

keys

1.19

Keys

1.08

 Keys

1.05

*/(

0.94

strings

0.87

stroke

0.86

notes

0.84

key

0.83

ilant

0.82

Activations Density 0.010%

keywords or phrases related to physical keys

mentions of "keys" in various contexts

No Comments

No Known Activations