INDEX

Explanations

mentions of notable people's names along with either positive or negative attributes

oai_token-act-pair · gpt-3.5-turbo

assertions about notable figures or their attributes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

urry

-0.72

ear

-0.61

âĵĺ

-0.59

Else

-0.59

IFF

-0.58

Cou

-0.57

âĶĢâĶĢ

-0.57

 occurs

-0.56

AGES

-0.56

ARM

-0.56

POSITIVE LOGITS

 indeed

1.06

 certainly

1.02

 undoubtedly

0.99

 undeniably

0.99

 notoriously

0.97

 fundamentally

0.93

 unquestion

0.90

 understandably

0.86

 profoundly

0.86

 hardly

0.85

Activations Density 0.562%

mentions of notable people's names along with either positive or negative attributes

assertions about notable figures or their attributes

No Comments

No Known Activations