INDEX

Explanations

instances of statements indicating negation or the absence of an expected outcome

oai_token-act-pair · gpt-3.5-turbo

expressions that indicate something that was not intended or could not happen

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

 antioxid

-0.77

 Tags

-0.77

:{

-0.71

onz

-0.67

achev

-0.67

ahime

-0.65

{"

-0.64

pour

-0.63

eva

-0.62

arling

-0.61

POSITIVE LOGITS

theless

1.12

 questioned

0.91

 mentioned

0.87

 truly

0.85

 prosecuted

0.83

 happier

0.80

entimes

0.80

 discussed

0.78

 remotely

0.77

 satisf

0.76

Activations Density 0.045%

instances of statements indicating negation or the absence of an expected outcome

expressions that indicate something that was not intended or could not happen

No Comments

No Known Activations