INDEX

Explanations

phrases related to protective cases or legal cases

oai_token-act-pair · gpt-3.5-turbo

instances of the term "case" to identify contexts where discussions or examples are being presented

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

etsk

-0.81

pez

-0.76

 destro

-0.74

 mathemat

-0.74

autical

-0.73

nesday

-0.71

¥ŀ

-0.71

slow

-0.69

uggish

-0.69

estern

-0.68

POSITIVE LOGITS

cases

0.98

cmp

0.79

case

0.79

backs

0.77

Kas

0.69

 Keen

0.68

suit

0.68

 Case

0.67

back

0.67

 WHEN

0.66

Activations Density 0.031%

phrases related to protective cases or legal cases

instances of the term "case" to identify contexts where discussions or examples are being presented

No Comments

No Known Activations