INDEX

Explanations

the word "unsub" or variations of it in the text

oai_token-act-pair · gpt-3.5-turbo

terminology related to subscription and unsubscribing from services or lists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 guiActiveUnfocused

-0.86

 Drawn

-0.74

iard

-0.74

 Corps

-0.74

OPLE

-0.69

 Mavericks

-0.69

 Drawing

-0.69

 Papers

-0.69

 Place

-0.68

RD

-0.68

POSITIVE LOGITS

scribe

1.16

 unsub

1.10

stant

1.09

mitted

1.06

tle

1.04

stantial

0.97

verting

0.95

comm

0.94

verse

0.90

scribed

0.90

Activations Density 0.008%

the word "unsub" or variations of it in the text

terminology related to subscription and unsubscribing from services or lists

No Comments

No Known Activations