INDEX

Explanations

proper names, specifically those with the name "Len"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Len."

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

xual

-0.77

 Typhoon

-0.71

 Occupations

-0.67

FUL

-0.64

AGES

-0.62

CLASSIFIED

-0.61

 Flavoring

-0.61

 Monstrous

-0.61

worldly

-0.60

 Seeking

-0.60

POSITIVE LOGITS

iency

1.04

ovo

1.01

uci

0.95

essee

0.92

seys

0.87

hart

0.87

opoly

0.86

oir

0.82

ikuman

0.82

unci

0.82

Activations Density 0.010%

proper names, specifically those with the name "Len"

occurrences of the name "Len."

No Comments

No Known Activations