INDEX

Explanations

mentions of colleagues or coworkers

oai_token-act-pair · gpt-3.5-turbo

references to 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 @ 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

aval

-0.80

 Ukrain

-0.71

amac

-0.71

finger

-0.70

sole

-0.69

uder

-0.68

si

-0.67

sword

-0.66

Clear

-0.65

hd

-0.64

POSITIVE LOGITS

 colleagues

1.12

 colleague

1.01

 coworkers

0.83

 classmates

0.74

rity

0.72

heses

0.70

ries

0.70

ratulations

0.70

opausal

0.68

utenant

0.67

Activations Density 0.022%

mentions of colleagues or coworkers

references to colleagues in various contexts

No Comments

No Known Activations