INDEX

Explanations

key terms related to media or entertainment, specifically focusing on titles or names

oai_token-act-pair · gpt-3.5-turbo

references to power dynamics in political 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

afety

-0.84

channelAvailability

-0.72

FINEST

-0.67

ascript

-0.66

azard

-0.65

 isEnabled

-0.64

anmar

-0.62

Topic

-0.60

hur

-0.60

 reins

-0.60

POSITIVE LOGITS

 Poker

0.76

ciating

0.72

 Hutch

0.67

 Females

0.66

Irish

0.66

 Sail

0.65

Sag

0.65

 Nemesis

0.63

 Skydragon

0.63

 Bucc

0.63

Activations Density 0.206%

key terms related to media or entertainment, specifically focusing on titles or names

references to power dynamics in political contexts

No Comments

No Known Activations