INDEX

Explanations

mentions of unsubstantiated claims or unsubscribing from mailing lists

oai_token-act-pair · gpt-3.5-turbo

terms related to unsubscribing or withdrawal from lists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 guiActiveUnfocused

-0.86

iard

-0.73

 Corps

-0.72

ordan

-0.70

 Wine

-0.69

 Palace

-0.69

 Elves

-0.68

 Pell

-0.68

SEA

-0.67

 Mavericks

-0.66

POSITIVE LOGITS

scribe

1.18

mitted

1.14

stant

1.12

 unsub

1.07

verting

1.01

scribed

0.99

tle

0.96

stantial

0.95

missible

0.94

verse

0.93

Activations Density 0.005%

mentions of unsubstantiated claims or unsubscribing from mailing lists

terms related to unsubscribing or withdrawal from lists

No Comments

No Known Activations