INDEX

Explanations

phrases or words related to a specific individual or entity, likely named "Der" with varying activations

oai_token-act-pair · gpt-3.5-turbo

references to the term "Der" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 taco

-0.72

 canoe

-0.66

TPS

-0.65

oka

-0.65

ãĥīãĥ©

-0.64

box

-0.64

ogg

-0.63

poon

-0.63

 ping

-0.62

omn

-0.62

POSITIVE LOGITS

Der

4.05

Der

3.16

der

1.65

der

1.38

 Derby

1.29

 dermat

1.29

 derivatives

1.25

 deriv

1.06

Die

1.05

 derivative

1.04

Activations Density 0.019%

phrases or words related to a specific individual or entity, likely named "Der" with varying activations

references to the term "Der" in various contexts

No Comments

No Known Activations