INDEX

Explanations

references to geographical locations, mainly villages or rural areas

oai_token-act-pair · gpt-3.5-turbo

phrases associated with areas or locations that are unaffected or have distinct characteristics

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

 Lifetime

-0.67

 omission

-0.57

isman

-0.52

aughs

-0.51

 formulation

-0.51

 cents

-0.50

neau

-0.49

Reference

-0.49

oyal

-0.49

raq

-0.49

POSITIVE LOGITS

 inhabited

0.79

Located

0.71

 habitable

0.69

 outskirts

0.69

near

0.68

 wetlands

0.67

 landmarks

0.66

 UNESCO

0.66

 encl

0.65

 inhabitants

0.65

Activations Density 1.231%

references to geographical locations, mainly villages or rural areas

phrases associated with areas or locations that are unaffected or have distinct characteristics

No Comments

No Known Activations