INDEX

Explanations

the name "Bundy"

oai_token-act-pair · gpt-3.5-turbo

mentions of the Bundy family and related entities or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ournal

-0.74

ORED

-0.65

akespeare

-0.65

oen

-0.65

ored

-0.64

alled

-0.64

erver

-0.64

raints

-0.64

ointed

-0.63

reads

-0.63

POSITIVE LOGITS

 Bundy

1.37

 ranch

0.90

anca

0.75

querque

0.74

 Ranch

0.74

anan

0.73

 ranc

0.71

bryce

0.70

ston

0.70

©¶æ¥µ

0.69

Activations Density 0.009%

the name "Bundy"

mentions of the Bundy family and related entities or events

No Comments

No Known Activations