INDEX

Explanations

mentions of specific locations or people's names

oai_token-act-pair · gpt-3.5-turbo

the word "ag" and its context in various usages or forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 reps

-0.67

 Widow

-0.63

 Bolt

-0.62

 Templ

-0.61

 bell

-0.60

 mileage

-0.59

 spinal

-0.58

LINE

-0.58

 Infinite

-0.58

wre

-0.57

POSITIVE LOGITS

ag

4.13

ags

2.35

AG

1.98

agos

1.86

agg

1.84

agging

1.73

agin

1.72

agged

1.67

agy

1.61

agn

1.59

Activations Density 0.017%

mentions of specific locations or people's names

the word "ag" and its context in various usages or forms

No Comments

No Known Activations