INDEX

Explanations

mentions of states or locations

oai_token-act-pair · gpt-3.5-turbo

references to "state" in various contexts

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

poons

-0.96

edin

-0.81

Maw

-0.78

 Aires

-0.74

azi

-0.74

 Miko

-0.68

ãĥīãĥ©

-0.67

lio

-0.67

bats

-0.67

ayson

-0.67

POSITIVE LOGITS

 legislatures

1.03

 legislature

0.86

rooms

0.84

 governments

0.83

 governors

0.81

wide

0.80

 legislators

0.76

less

0.74

SPONSORED

0.74

 Legislature

0.70

Activations Density 0.063%

mentions of states or locations

references to "state" in various contexts

No Comments

No Known Activations