INDEX

Explanations

mentions of a specific entity or organization ("Ars") within a document

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations or entities

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

 Cruise

-0.71

woods

-0.69

 cruise

-0.68

mouth

-0.66

plant

-0.65

 Survivor

-0.65

olls

-0.64

 abortion

-0.64

park

-0.63

 sett

-0.63

POSITIVE LOGITS

Ars

3.59

EFF

1.38

EFF

1.12

 IEEE

1.02

Tech

1.01

Ess

0.94

Bloomberg

0.92

ertodd

0.90

 Sebast

0.89

 Dahl

0.89

Activations Density 0.021%

mentions of a specific entity or organization ("Ars") within a document

references to specific organizations or entities

No Comments

No Known Activations