INDEX

Explanations

proper nouns related to politics and individuals

oai_token-act-pair · gpt-3.5-turbo

consonant clusters or specific letter patterns in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 conclud

-0.69

 skelet

-0.67

 carbohyd

-0.64

ãĥ¼ãĥĨ

-0.63

CONCLUS

-0.61

 encount

-0.61

BOOK

-0.61

 conduc

-0.61

 ACTIONS

-0.60

 occas

-0.55

POSITIVE LOGITS

oda

0.78

anski

0.77

amer

0.76

oe

0.74

uk

0.72

nik

0.71

la

0.70

ilia

0.68

anders

0.68

onga

0.67

Activations Density 0.364%

proper nouns related to politics and individuals

consonant clusters or specific letter patterns in text

No Comments

No Known Activations