INDEX

Explanations

names and terms related to Sherlock Holmes and his creator, Arthur Conan Doyle

oai_token-act-pair · gpt-3.5-turbo

mentions of the character Sherlock Holmes and related names

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

issan

-0.84

auri

-0.83

ensing

-0.78

isconsin

-0.75

ancies

-0.72

acid

-0.72

urt

-0.71

nir

-0.70

ouver

-0.70

esa

-0.70

POSITIVE LOGITS

 Holmes

1.25

 Sherlock

1.04

 Doyle

0.87

Sher

0.82

 Reborn

0.76

batch

0.73

 Cumber

0.73

Lens

0.71

 Underground

0.69

Girls

0.67

Activations Density 0.022%

names and terms related to Sherlock Holmes and his creator, Arthur Conan Doyle

mentions of the character Sherlock Holmes and related names

No Comments

No Known Activations