INDEX

Explanations

proper nouns followed by location identifiers

oai_token-act-pair · gpt-3.5-turbo

references to locations associated with organizations, particularly in a professional context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

osen

-0.73

-0.72

Balt

-0.71

osure

-0.68

ests

-0.66

saf

-0.65

older

-0.65

dimension

-0.65

estern

-0.64

 Bagg

-0.64

POSITIVE LOGITS

quartered

1.09

 operator

0.77

 Pixie

0.70

 entrepreneur

0.68

 activist

0.68

 newcom

0.67

 nonprofit

0.67

 headquartered

0.66

 worker

0.66

venture

0.65

Activations Density 0.024%

proper nouns followed by location identifiers

references to locations associated with organizations, particularly in a professional context

No Comments

No Known Activations