INDEX

Explanations

phrases related to official confirmations or statements

oai_token-act-pair · gpt-3.5-turbo

mentions of spokespersons and their statements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uterte

-0.96

duino

-0.89

terness

-0.86

umbledore

-0.84

obyl

-0.81

ascus

-0.80

 dstg

-0.80

paio

-0.79

ettel

-0.79

haust

-0.79

POSITIVE LOGITS

 company

1.09

Company

1.02

company

0.92

 Company

0.87

 subsidiary

0.82

quartered

0.80

CEO

0.80

 Website

0.79

 subsidiaries

0.77

 organization

0.77

Activations Density 0.934%

phrases related to official confirmations or statements

mentions of spokespersons and their statements

No Comments

No Known Activations