INDEX

Explanations

words related to workplace interactions, specifically interactions with coworkers

oai_token-act-pair · gpt-3.5-turbo

references to coworkers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ciating

-0.69

rawn

-0.65

itars

-0.63

 original

-0.60

 heaviest

-0.59

hran

-0.57

sembly

-0.57

eq

-0.56

si

-0.55

 unusual

-0.53

POSITIVE LOGITS

jee

0.82

berus

0.76

 laun

0.76

rior

0.76

minster

0.74

adish

0.73

nel

0.70

rer

0.70

Tube

0.69

naire

0.69

Activations Density 0.057%

words related to workplace interactions, specifically interactions with coworkers

references to coworkers

No Comments

No Known Activations