INDEX

Explanations

names or terms related to a character called "Herbie."

oai_token-act-pair · gpt-3.5-turbo

the term "bie" and its variations, likely referring to a type of identity or classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 suspense

-0.77

 fixation

-0.70

 perpetual

-0.70

 tense

-0.70

 pollutants

-0.68

 totality

-0.68

 unavoidable

-0.67

Dialogue

-0.66

 polarized

-0.66

orses

-0.66

POSITIVE LOGITS

bie

1.23

gee

1.12

yip

0.95

gie

0.91

bies

0.89

GES

0.89

gment

0.86

lly

0.85

leck

0.83

butt

0.83

Activations Density 0.008%

names or terms related to a character called "Herbie."

the term "bie" and its variations, likely referring to a type of identity or classification

No Comments

No Known Activations