INDEX

Explanations

mentions of friends or friendly interactions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "friend"

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

hner

-0.77

arcity

-0.74

inth

-0.71

interstitial

-0.70

ammers

-0.68

authorized

-0.65

etheus

-0.64

argon

-0.61

tan

-0.61

intensity

-0.61

POSITIVE LOGITS

lier

1.06

friend

1.06

 friend

1.05

 buddy

1.01

 friends

1.00

liness

0.97

liest

0.97

 buddies

0.96

 Ally

0.90

Friends

0.88

Activations Density 0.016%

mentions of friends or friendly interactions

instances of the word "friend"

No Comments

No Known Activations