INDEX

Explanations

instances of the word "obvious"

oai_token-act-pair · gpt-3.5-turbo

the word "obvious" and its variants, indicating a focus on clear or evident statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nan

-0.91

ingers

-0.77

iership

-0.73

borg

-0.73

rams

-0.72

rigan

-0.70

isol

-0.68

psey

-0.68

restling

-0.67

uden

-0.66

POSITIVE LOGITS

iary

1.00

 obvious

0.83

 Leilan

0.80

 signs

0.76

 contrad

0.73

 resemblance

0.73

 distinction

0.72

 Signs

0.70

ances

0.70

\\\\\\\\

0.70

Activations Density 0.012%

instances of the word "obvious"

the word "obvious" and its variants, indicating a focus on clear or evident statements

No Comments

No Known Activations