INDEX

Explanations

dialogue segments encapsulated in quotation marks

oai_token-act-pair · gpt-3.5-turbo

repeated quotation marks in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 fres

-0.79

 prec

-0.74

 sprint

-0.73

 endeav

-0.72

 valued

-0.72

 adjud

-0.71

 care

-0.70

spr

-0.70

 repro

-0.70

pir

-0.70

POSITIVE LOGITS

Obviously

1.25

Certainly

1.25

That

1.25

Because

1.25

What

1.24

But

1.24

We

1.23

Indeed

1.22

They

1.22

Yeah

1.22

Activations Density 0.119%

dialogue segments encapsulated in quotation marks

repeated quotation marks in the text

No Comments

No Known Activations