INDEX

Explanations

mentions of fictional characters

oai_token-act-pair · gpt-3.5-turbo

references to specific characters 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 @ 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

hower

-0.74

rup

-0.71

VERTIS

-0.71

yg

-0.70

atory

-0.69

atories

-0.68

isco

-0.67

liga

-0.65

ledge

-0.64

gard

-0.62

POSITIVE LOGITS

acters

1.23

istically

1.05

istics

0.97

izations

0.93

 assassinate

0.85

 arcs

0.84

isations

0.81

izes

0.80

 actor

0.80

 assassination

0.77

Activations Density 0.021%

mentions of fictional characters

references to specific characters in various contexts

No Comments

No Known Activations