INDEX

Explanations

phrases indicating doubt or uncertainty

oai_token-act-pair · gpt-3.5-turbo

instances of the word "would."

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

DN

-0.62

Kag

-0.61

 Case

-0.61

Kis

-0.61

âĢ¢âĢ¢âĢ¢âĢ¢

-0.61

notes

-0.60

Building

-0.59

SIG

-0.59

 marker

-0.57

DOC

-0.57

POSITIVE LOGITS

be

1.01

 prefer

0.94

nt

0.89

 qualify

0.86

 tolerate

0.86

 gladly

0.84

 allow

0.83

itate

0.83

 consider

0.83

 suffice

0.82

Activations Density 0.149%

phrases indicating doubt or uncertainty

instances of the word "would."

No Comments

No Known Activations