INDEX

Explanations

words related to companionship or companions

oai_token-act-pair · gpt-3.5-turbo

references to companions, highlighting both animate and inanimate associates

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

©¶æ

-0.74

uman

-0.73

UGE

-0.73

uden

-0.67

uge

-0.67

ihil

-0.66

mberg

-0.66

prus

-0.65

ramid

-0.64

maxwell

-0.64

POSITIVE LOGITS

 companions

0.99

 companion

0.99

 Companion

0.82

":[{"

0.78

atives

0.77

ively

0.76

hesis

0.74

compan

0.73

mented

0.72

ments

0.71

Activations Density 0.014%

words related to companionship or companions

references to companions, highlighting both animate and inanimate associates

No Comments

No Known Activations