INDEX

Explanations

questions or inquiry-related phrases

oai_token-act-pair · gpt-3.5-turbo

rhetorical questions or phrases emphasizing inquiry

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

shore

-0.66

roads

-0.61

trop

-0.60

eer

-0.59

Gy

-0.59

idon

-0.59

println

-0.59

ulic

-0.58

ped

-0.58

gal

-0.57

POSITIVE LOGITS

soever

1.27

 happens

1.12

 happened

1.04

 transpired

0.97

 distinguishes

0.96

 happ

0.90

 else

0.85

 ensued

0.84

 constitutes

0.83

 separates

0.83

Activations Density 0.086%

questions or inquiry-related phrases

rhetorical questions or phrases emphasizing inquiry

No Comments

No Known Activations