INDEX

Explanations

phrases indicating logical reasoning or justification

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "makes sense" or similar expressions indicating reasoning or logic

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

opers

-0.69

 downed

-0.61

 Tempest

-0.60

 Spectre

-0.59

enne

-0.59

anche

-0.56

chu

-0.56

 laun

-0.55

les

-0.55

available

-0.54

POSITIVE LOGITS

 logically

1.08

why

0.96

WHY

0.86

 given

0.85

 considering

0.84

 intuitive

0.83

why

0.80

 financially

0.78

 economically

0.74

:{

0.72

Activations Density 0.055%

phrases indicating logical reasoning or justification

instances of the phrase "makes sense" or similar expressions indicating reasoning or logic

No Comments

No Known Activations