INDEX

Explanations

the word "loser" or words containing "loser"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "loser" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 shroud

-0.81

canon

-0.76

 Blessing

-0.72

 Catalyst

-0.70

Nir

-0.68

opathy

-0.66

Acknowled

-0.65

 Aurora

-0.65

 prescribing

-0.65

Bio

-0.65

POSITIVE LOGITS

 losers

2.70

 loser

2.66

 contestants

1.25

 contestant

1.19

 underdog

1.17

 liar

1.10

onge

1.08

 elimination

1.06

 improv

1.05

Redditor

1.01

Activations Density 0.044%

the word "loser" or words containing "loser"

instances of the word "loser" and its variations

No Comments

No Known Activations