INDEX

Explanations

proper nouns related to literature and media

oai_token-act-pair · gpt-3.5-turbo

references to notable individuals and cultural works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hips

-0.84

cheat

-0.77

ees

-0.76

arnaev

-0.70

lies

-0.70

Domin

-0.70

irmed

-0.69

anova

-0.69

iege

-0.68

een

-0.67

POSITIVE LOGITS

 Simpson

0.87

ufact

0.79

 pitched

0.72

 Homer

0.70

 Kurd

0.69

 Gohan

0.69

ophon

0.67

Hes

0.66

icably

0.63

ocrates

0.60

Activations Density 0.008%

proper nouns related to literature and media

references to notable individuals and cultural works

No Comments

No Known Activations