INDEX

Explanations

phrases related to making arguments or presenting points

oai_token-act-pair · gpt-3.5-turbo

instances of the word "argument" and related discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ookie

-0.66

 livest

-0.64

 Seym

-0.64

ISTER

-0.63

 Atomic

-0.62

 Carbuncle

-0.61

eco

-0.61

cler

-0.61

 curfew

-0.61

 PHOTO

-0.60

POSITIVE LOGITS

ative

1.24

uments

1.16

 against

1.04

 arguments

0.95

abl

0.94

ument

0.93

 argument

0.86

 arguing

0.85

 persu

0.84

 Against

0.84

Activations Density 0.039%

phrases related to making arguments or presenting points

instances of the word "argument" and related discussions

No Comments

No Known Activations