INDEX

Explanations

dialogue excerpts containing quotes or speaking actions

oai_token-act-pair · gpt-3.5-turbo

dialogue or conversational lines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

etheless

-0.93

 coerc

-0.93

 contrasts

-0.84

 incentives

-0.84

 exploited

-0.81

 compuls

-0.80

 comprehens

-0.79

 clauses

-0.77

 merciless

-0.77

ilitarian

-0.77

POSITIVE LOGITS

Hello

1.62

Hi

1.51

 Hello

1.40

 hello

1.39

Hi

1.35

Hey

1.34

hello

1.30

Yo

1.20

hi

1.20

Hey

1.19

Activations Density 0.200%

dialogue excerpts containing quotes or speaking actions

dialogue or conversational lines

No Comments

No Known Activations