INDEX

Explanations

terms related to corporations or companies

oai_token-act-pair · gpt-3.5-turbo

instances of the word "corp" and its variations

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

lihood

-0.75

TPS

-0.70

 Hits

-0.70

 Finder

-0.68

HAHAHAHA

-0.66

éĹĺ

-0.66

CRIP

-0.66

 aloud

-0.64

STER

-0.64

HUD

-0.63

POSITIVE LOGITS

porate

1.45

ruption

1.27

poral

1.23

ridor

1.23

rupted

1.22

rection

1.18

rupt

1.16

respond

1.14

por

1.10

pse

1.09

Activations Density 0.011%

terms related to corporations or companies

instances of the word "corp" and its variations

No Comments

No Known Activations