INDEX

Explanations

mentions of different 'voices' or viewpoints in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to diverse perspectives and expressions

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

rip

-0.72

photos

-0.69

artment

-0.67

 postp

-0.66

tenance

-0.66

armac

-0.65

addy

-0.64

ns

-0.64

nex

-0.62

xon

-0.62

POSITIVE LOGITS

 voices

1.36

voice

1.23

 voice

1.00

 Voices

0.91

 vocals

0.89

 voiced

0.87

 louder

0.87

 voic

0.87

 singers

0.86

linger

0.84

Activations Density 0.011%

mentions of different 'voices' or viewpoints in various contexts

references to diverse perspectives and expressions

No Comments

No Known Activations