INDEX

Explanations

questions

oai_token-act-pair · gpt-3.5-turbo

questions 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 tremend

-0.69

 princ

-0.67

 scrut

-0.66

bidden

-0.66

 civilisation

-0.65

 strugg

-0.65

 unbeliev

-0.65

 reckoning

-0.65

 notor

-0.64

uckland

-0.64

POSITIVE LOGITS

 Were

1.35

 Explain

1.29

Did

1.22

Was

1.22

 What

1.22

 Does

1.19

How

1.19

 Would

1.17

Are

1.17

 Where

1.16

Activations Density 0.079%

questions

questions in the text

No Comments

No Known Activations