INDEX

Explanations

mentions of a specific entity named "Dog" in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of "Dog" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Mb

-0.70

lished

-0.69

 unders

-0.68

 exch

-0.66

olate

-0.66

 GOODMAN

-0.66

 unres

-0.63

 allowances

-0.63

CLOSE

-0.62

olated

-0.61

POSITIVE LOGITS

Dog

3.86

Dog

2.85

 Dogs

2.51

dog

2.12

dog

2.11

 dogs

1.91

 canine

1.58

 Pupp

1.46

 Animal

1.42

dogs

1.41

Activations Density 0.012%

mentions of a specific entity named "Dog" in various contexts

mentions of "Dog" in various contexts

No Comments

No Known Activations