INDEX

Explanations

locations or directional terms within a region or city

oai_token-act-pair · gpt-3.5-turbo

references to geographical locations and their parts in a spatial context

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

Bir

-0.79

ÃįÃį

-0.75

PLIED

-0.71

ado

-0.70

oaded

-0.70

sense

-0.66

Lear

-0.65

GM

-0.64

atonin

-0.63

Choice

-0.63

POSITIVE LOGITS

 thereof

0.82

 ridge

0.77

of

0.71

 corner

0.70

 latitude

0.70

 mainline

0.69

 plateau

0.67

 inhabited

0.66

 tract

0.66

 Hispan

0.66

Activations Density 0.135%

locations or directional terms within a region or city

references to geographical locations and their parts in a spatial context

No Comments

No Known Activations