INDEX

Explanations

mentions of colleagues or coworkers

oai_token-act-pair · gpt-3.5-turbo

the term "colleagues" 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 @ 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

aval

-0.79

 Ukrain

-0.70

ball

-0.68

si

-0.65

uder

-0.65

iculture

-0.64

amac

-0.64

istic

-0.63

forest

-0.63

direction

-0.62

POSITIVE LOGITS

 colleagues

1.11

 colleague

1.03

 coworkers

0.86

ority

0.74

ratulations

0.71

 classmates

0.71

utenant

0.70

 roomm

0.68

 cowork

0.67

 tasked

0.67

Activations Density 0.021%

mentions of colleagues or coworkers

the term "colleagues" in various contexts

No Comments

No Known Activations