INDEX

Explanations

proper nouns related to a specific individual named Nielsen

oai_token-act-pair · gpt-3.5-turbo

references to individuals with the surname 'Nielsen'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

course

-0.75

 Carbuncle

-0.73

ebook

-0.71

scene

-0.67

 Izan

-0.65

draw

-0.64

fare

-0.64

xual

-0.62

 Continental

-0.61

 Icar

-0.61

POSITIVE LOGITS

elsen

1.54

 Nielsen

1.51

acher

0.88

elson

0.79

 "$:/

0.79

achers

0.79

ylon

0.77

ihilation

0.74

obook

0.73

 Rasmussen

0.73

Activations Density 0.005%

proper nouns related to a specific individual named Nielsen

references to individuals with the surname 'Nielsen'

No Comments

No Known Activations