INDEX

Explanations

references to the name "Lin" with varying degrees of specificity

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual named Lin

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

 Seym

-1.02

ĳå£«

-0.81

sburgh

-0.76

 therap

-0.72

GROUND

-0.71

theless

-0.70

lain

-0.67

ledged

-0.67

FUL

-0.66

EStream

-0.65

POSITIVE LOGITS

eman

0.95

ergy

0.94

emate

0.92

ergic

0.91

otype

0.90

sey

0.88

nea

0.86

jiang

0.85

olen

0.85

iets

0.83

Activations Density 0.011%

references to the name "Lin" with varying degrees of specificity

references to a specific individual named Lin

No Comments

No Known Activations