INDEX

Explanations

mentions of family members, specifically sons

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "son."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

veyard

-0.79

 hazards

-0.65

Union

-0.64

 manifold

-0.61

iculty

-0.60

Population

-0.60

 freight

-0.60

PORT

-0.59

 accommodations

-0.58

 mechanisms

-0.57

POSITIVE LOGITS

ogram

1.03

nets

1.00

hood

0.95

 Barron

0.89

 Gohan

0.87

pins

0.85

hesis

0.84

ograms

0.83

son

0.81

zai

0.81

Activations Density 0.032%

mentions of family members, specifically sons

mentions of the word "son."

No Comments

No Known Activations