INDEX

Explanations

phrases related to two-part actions or descriptions

oai_token-act-pair · gpt-3.5-turbo

references to political systems or frameworks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Lanka

-0.64

AVG

-0.63

 Claus

-0.63

 Gillespie

-0.63

 Bland

-0.61

CoC

-0.61

 Kung

-0.60

 Canaver

-0.60

 Richards

-0.60

POSITIVE LOGITS

thirds

1.79

dimensional

1.50

legged

1.40

dozen

1.35

hander

1.33

fif

1.33

year

1.32

hour

1.32

minute

1.31

tier

1.31

Activations Density 0.044%

phrases related to two-part actions or descriptions

references to political systems or frameworks

No Comments

No Known Activations