INDEX

Explanations

names of people, specifically focusing on a particular surname

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "aren" in the text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Labrador

-0.70

 jerk

-0.69

 extent

-0.66

 srfAttach

-0.64

Fish

-0.61

 degree

-0.59

 measure

-0.58

 Kinnikuman

-0.58

 magnitude

-0.58

 Accuracy

-0.57

POSITIVE LOGITS

aren

1.08

wen

0.91

isma

0.89

furt

0.88

stant

0.87

awan

0.84

sten

0.82

zen

0.81

wyn

0.81

wana

0.81

Activations Density 0.008%

names of people, specifically focusing on a particular surname

occurrences of the term "aren" in the text

No Comments

No Known Activations