INDEX

Explanations

terms related to voicemail and email

oai_token-act-pair · gpt-3.5-turbo

mentions of voicemail communication

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ĭ±

-0.67

Jiu

-0.64

ablishment

-0.62

abama

-0.61

Bat

-0.59

 guiActiveUnfocused

-0.59

 Nanto

-0.59

å°Ĩ

-0.59

abad

-0.58

hov

-0.57

POSITIVE LOGITS

 voic

1.54

email

1.17

etooth

1.15

ework

1.02

eless

1.01

elines

0.88

 chords

0.84

eware

0.79

iance

0.78

lyak

0.77

Activations Density 0.005%

terms related to voicemail and email

mentions of voicemail communication

No Comments

No Known Activations